Reference – Graphics – Dissertation – Methodology – Researches – Countries – Age – Genetics – Human Expectancy of Life – Contribuition to the world scientific community and other people worldwide – Fact – Reality % Mouse – Probabilities @ Work – Innovation – Prizes – Investments & Laboratories – Mice and Human Researches – Statistics – Epidemiology – Countries – Diseases – Health @-> Links and videos about mice researches – Images – Time – Statistics @ 12 Innovations That Will Change Health Care and Medicine in the 2020s % New ‘prime’ genome editor could surpass CRISPR & A Guide To Machine Learning Interview Questions And Answers @ ´´In patent law the citation of previous works, or prior art, helps establish the uniqueness of the invention being described. Modern scientists are sometimes judged by the number of times their work is cited by others—this is actually a key indicator of the relative importance of a work in science.´´ @ https://en.wikipedia.org/wiki/Scientific_citation @ Other Very Important Facts

https://en.wikipedia.org/wiki/Scientific_citation

@ NASA shares a spooky photo of the sun looking like a giant flaming jack-o’-lantern @ SpaceX reveals early users of satellite-based high-speed internet & New gene editing technology could correct 89% of genetic defects # CRISPR GENE EDITING – Top Machine Learning Interview Questions You Must Prepare In 2019

https://www.edureka.co/blog/interview-questions/machine-learning-interview-questions/

https://edition.cnn.com/2019/10/22/health/gene-editing-study-intl-hnk-scli-scn/index.html?fbclid=IwAR0gP2XVvqs4I9Qo_wFGevPPm3GBr6_xZvVQj5HuvpdyIZPuhA5_LXVGB74

https://medium.com/edureka/machine-learning-interview-questions-a5aef8a3ca60

https://www.sciencemag.org/news/2019/10/new-prime-genome-editor-could-surpass-crispr?fbclid=IwAR2PCQJRCVCh9PsF-lKv-Myy9xgX7pxvcbjdzoKZOByQngijeOnvL4qYqMc

The European Animal Research Association (EARA) -> http://eara.eu/en/

http://www.animalresearch.info/en/designing-research/research-animals/mouse/ https://www.the-scientist.com/tag/mouse-research

https://www.the-scientist.com/tag/mouse-research

https://www.kentscientific.com/blog/mice-vs-rats-in-research-whats-the-difference/

https://en.wikipedia.org/wiki/Laboratory_mouse

http://www.understandinganimalresearch.org.uk/animals/10-facts/mouse/

https://www.yourgenome.org/facts/why-use-the-mouse-in-research

https://www.jax.org/why-the-mouse

https://www.livescience.com/32860-why-do-medical-researchers-use-mice.html

Fighting cancer: Animal research at Cambridge 37.457 visualizações•24 de abr. de 2015 Cambridge University 175 mil inscritos
´´Animal research plays an essential role in our understanding of health and disease and in the development of modern medicine and surgical techniques. As part of our commitment to openness, this film examines how mice are helping the fight against cancer. It takes a in-depth look at the facilities in which they are housed, exploring issues of animal welfare and the search for replacements.
We welcome comments about this article. However, as with discussions on all of our news and feature pages, comments will be moderated so please do not post contributions that are offensive or contain profanities and do stay on topic. We do not moderate comments in real-time so do bear with us if there is a delay before they appear.´´

Categoria

Ciência e tecnologia

Why mice are the best candidates for research.

7.034 visualizações•30 de out. de 2017
Inside Science
15 mil inscritos (Inside Science) — ´´The mouse … this tiny creature has had a huge impact on science research. Mice make excellent models for human disease because parts of their DNA are similar to human DNA. They suffer from many of the same ailments as humans, such as cancer, diabetes and even anxiety. Research on mice has helped scientists understand the causes of many different diseases — and the use of brain imaging with mice has been a big boost to studies of brain function in humans. But mice brains are tiny, and guess what? Imaging a little brain for research isn’t easy. Now scientists may have found a better, easier way to help further research. To some people, this might be considered a pesky pest, but to Adam Bauer, at the Washington University School of Medicine in St. Louis, this little creature is an amazing research tool. “The beauty of imaging mouse models or even rat models or monkeys or people is that all of us, all mammals have hemoglobin in their brains,” said Bauer. Hemoglobin is the molecule in blood that carries oxygen. It makes it easier to take an image of the brain because it works as a good contrast agent during a functional MRI scan — helping to light up certain parts of the brain. “That’s something that we can all image, from humans all the way down to mice. And the fact that mice exhibit these functional networks that resemble very closely to what we see in people is really powerful,” said Bauer. Just like doctors use functional MRI — or FMRI — to look at human brain activity, Bauer uses a similar technique — but for tiny mouse brain activity. The technology shows which area of the brain is in use during a particular mental process while a mouse is awake. “What we develop is kind of an FMRI surrogate for use in mice. So, up until recently, FMRI was having a pretty difficult time creating really nice images in the mouse brain using the standard techniques that they use in people. It’s come a long way and they’re able to generate some really beautiful images now, looking at changes in the blood activity in the brain,” said Bauer. “The mice will be brought into our imaging systems and we’ll image their activity, they’ll be very specific behaviors that we will have these mice perform or specific tasks that we’ll have these mice perform. So, certain regions of the brain are responsible for the execution of specific tasks,” said Bauer. Bauer also looks at the brains of mice that have been given a stroke. Researchers can learn how the mouse brain works to restore functional areas and remap itself to compensate for the damage from a stroke — and translate those results for restoring more function in humans. “And so, by imaging mice, you can learn a lot about human disease and how to affect it and treat it and make it better in people,” concluded Bauer. ´´

 

Using mice in hearing research

2.180 visualizações•5 de mar. de 2014
Medical Research Council
12,6 mil inscritos ´´How are mice helping with hearing research? Professor Steve Brown (http://www.insight.mrc.ac.uk/2012/05/…), the Director of the MRC Mammalian Genetics Unit (http://www.har.mrc.ac.uk/), carries out research investigating the genetic basis of deafness by changing specific genes in mice to find out their role in hearing. His work has led to the identification of a potential new treatment for glue ear (http://www.mrc.ac.uk/Newspublications…) in children. Here he describes the work of the MRC Mary Lyon Centre and shows us a hearing experiment in which an anaesthetised mouse is tested for its response to a particular tone. ´´

Hi, how are you? I hope you are well. Thank you for visiting my blog. I did it with so much dedication. This blog sharing is very important to scientific community as well as other people worlwide. My blog goal is help anyone in anyway who visit it through very updated, important and interesting health news. There are several links too. Note: I do not earn money from it. I never earner money with social networkings, blogs I did and other things.

Many people of different countries have visited my blog. There are 776 posts made by me, 72 followers and 3.4 thousand comments approved by me. I started to elaborate this blog in 2018.

I hope with it to contribute significantly to the technical-scientific and socio-economic development of the countries. The human expectancy of life worldwide needs to increase so much faster more and more urgently with very efficient and innovative researches.

My blog content www.science1984.wordpress.com is very good with high quality. There are very important informations for human health like my dissertation [The influence of physical activity in the progression of experimental lung cancer in mice – Pathol Res Pract. 2012 Jul 15;208(7):377-8] and my monograph (Induction of benzonidazole resistance in human isolates of Trypanosoma cruzi).

I did very detailed, innovative and important graphics about variations of all mice weights during all experimental time in my dissertation that are not in the scientific article as well as details about time of exercises and rest of the animals. They are very innovative graphics for the world scientific community. They can be an excellent reference for researches of many types like genetic engineering. Many people gave me positive feedbacks about it. These data are in my blog.

The age of the mouse and the human being with the genetics influence in certain ways in pathophysiology in the humans and mice. So, mice researches are very important for the society as well as researches with humans.

Very Important Observations:   

1. Cancer is very related to the weight loss of the patient. Weight loss of the patient is very associated with cancer – The syndrome of Anorexia-Cachexia (SAC) is a frequent complication in patients with advanced malignant neoplasia.

2. Age, weight and genetics of the person are very important factors that influence cancer in a determined ways.

3. The genetics of the mouse is very similar to that of the human.

4. Maintaining proper body weight is one of the main ways to prevent cancer of a person.

5. Animal testing has a very high importance to world society.

6. The mouse is the main animal model used as the basis for research on diseases that affect humans.

7. Weight lifting (bodybuilder) is a very good example of anaerobic physical activity in humans.

Many laboratories have been researching mice for a long time, even resulting in prizes for researchers such as the Nobel Prize. For example, the Jackson Laboratory.

There are in my blog data about the invitations I received by direct messages through the Internet to participate in 68 very important events in 31 cities in less than 2 years (Auckland, Melbourne, Toronto, Edinburgh, Madrid, Suzhou, Stanbul, Miami, Singapore, Kuala Lumpur, Abu Dhabi, San Diego, Bangkok, Dublin, Sao Paulo, Dubai, Boston, Berlin, Stockholm, Prague, Valencia, Osaka, Amsterdam, Helsinki, Paris, Tokyo, Vienna, Rome, Zurich, London and Frankfurt) because I participated of very important researches in Brazil.

In my dissertation the progression of lung cancer was lower in the group of mice that practiced anaerobic physical activity. It would be very innovative, important and interesting to do researches in mice and humans testing a substance or substances, analyzing the influences of ages, sex, genetics, weights and types of physical activities within the group itself and in the other groups in the inhibition and progression of cancer and in other situations in all experimental time, for example, and examining biochemical, pathological, pharmacological and physiological factors. In this context, it is essential to seek new methodologies for the treatment, prevention and early detection of cancer and other diseases, such as vaccines and other very modern technologies. It´s fundamental to consider the significance of variants like weights, ages and genetics in relation to cancer. It is not easy to understand cancer in all aspects. So, more researches about it are very necessary in the world.

Link of my dissertation: https://science1984.wordpress.com/2018/07/15/i-did-very-important-detailed-and-innovative-graphics-about-variations-of-all-mice-weigths-during-all-exerimental-time-my-dissertation-they-can-be-an-excelent-reference-for-future-researches-like-2/

Link of my monograph: https://science1984.wordpress.com/2018/07/15/my-monography-chagas-disease-research-in-laboratory-2/

I hope that researchers, teachers, students, scientists and other people linked to scientific researches worldwide use the graphics I made about the variations of all mice weights during all experimental time of my dissertation as an example, model or reference for conducting scientific researches as well as other data from my monograph and dissertation, leading to very beneficial innovations in the methodologies bringing very important and relevant results to the world society, significantly increasing human lifespan through very innovative and effective scientific researches.

Espero que os pesquisadores, professores, estudantes, cientistas e demais pessoas vinculadas no campo de pesquisas científicas no mundo todo utilizem os gráficos que fiz sobre as variações dos pesos de todos os camundongos durante todo o tempo experimental da minha dissertação como um exemplo, modelo ou referência para a realização de pesquisas científicas assim como outros dados da minha monografia e dissertação, fazendo com que haja uma inovação muito benéfica nas metodologias trazendo resultados bem importantes para a sociedade mundial, aumentando significativamente o tempo de vida humano cada vez mais por meio de pesquisas científicas muito inovadoras e eficazes.

Gratitude: I am very grateful because I was invited by Internet through direct messages to participate in 55 very important science events in the world in 25 cities in less than 1 year.

Visit, watch and share it if possible.

Thank you in advance.

Curriculum Lattes: http://buscatextual.cnpq.br/buscatextual/visualizacv.do?id=K4240145A2 

https://www.escavador.com/sobre/4321901/rodrigo-nunes-cal

https://www.researchgate.net/publication/225286318_The_influence_of_physical_activity_in_the_progression_of_experimental_lung_cancer_in_mice

The influence of physical activity in the progression of experimental lung cancer in mice -> Pathol Res Pract. 2012 Jul 15;208(7):377-81. doi: 10.1016/j.prp.2012.04.006. Epub 2012 Jun 8. https://www.ncbi.nlm.nih.gov/pubmed/22683274

Best wishes,

Sincerely,

Rodrigo Nunes Cal

  In: FACTSAnimals and Plants

Why use the mouse in research?

Humans and mice share many common genetic features and by examining the physiology, anatomy and metabolism of a mouse, scientists can gain a valuable insight into how humans function. 

Key facts

  • Over the past century, the house mouse (Mus musculus) has become the preferred mammalian model? for genetic research.
  • In the early days of biomedical research, scientists developed mouse models by selecting and breeding specific mice to produce offspring with certain desired characteristics.
  • Now scientists use mice to simulate human genetic disorders? in order to study their development and test new therapies.
  • As a scientific tool, mice have helped to speed up the progress of research and enabled the development of important new drugs?.
  • The genome? sequence of the mouse was published in December 2002.
  • Its genome is approximately 3,500 million base pairs? in length and contains over 23,000 protein-coding genes? (Ensembl).

An adult black mouse.

Image credit: Wellcome Library, London

Benefits of the mouse

  • The mouse has many similarities to humans in terms of anatomy, physiology and genetics.
  • The mouse genome is very similar to our own, making mouse genetic research particularly useful for the study of human diseases.
  • Mice are cost effective because they are cheap and easy to look after.
  • Adult mice multiply quickly. They can reproduce as often as every three weeks (they mate on the day they give birth), so scientists have lots of mice to work with.
  • The mouse is small, so convenient to house.
  • The time between a mouse being born and giving birth (generation time) is short, usually around 10 weeks. This means several generations can be observed at once.
  • The mouse has a short lifespan (one mouse year equals about 30 human years) which means scientists can easily measure the effects of ageing.
  • Mice are extremely useful for studying complex diseases?, such as atherosclerosis and hypertension, as many of the genes responsible for these diseases are shared between mice and humans. Research in mice provides insights into the genetic risk factors for these diseases in the human population.
  • It is relatively easy to manipulate the mouse genome, for example, adding or removing a gene to better understand its role in the body. This provides a powerful tool for modelling specific diseases when a mutated gene is known to play a role in the disease.
  • Mice are far better than flies or worms for studying complex biological systems found in humans, such as the immune, endocrine (delivers hormones? into the body), nervous, cardiovascular and skeletal systems. Like humans, mice naturally develop diseases that affect these systems, including cancer? and diabetes?.
  • Immunodeficient mice (mice without a fully functioning immune system) can also be used as hosts to grow both normal and diseased human tissue. This has been a useful tool in cancer? and AIDS? research. 

Is this page helpful?YesNoShareTweetPin

This page was last updated on 2017-03-03Of mice and menSTORIESThe mouse is closely related to humans with a striking similarity to us in terms of anatomy, physiology and genetics. This makes the mouse an extremely useful model organism. What are model organisms?FACTSA model organism is a species that has been widely studied, usually because it is easy to maintain and breed in a laboratory setting and has particular experimental advantages.Why use the zebrafish in research?FACTSSince the 1960s, the zebrafish (Danio rerio) has become increasingly important to scientific research. It has many characteristics that make it a valuable model for studying human genetics and disease.Why use the frog in research?FACTSMuch of our current knowledge about the mechanisms of early development in vertebrates comes from studies using the African clawed frog (Xenopus laevis) and Western clawed frog (Xenopus tropicalis).Why use the worm in research?FACTSWhile the fruit fly has a long history as a model organism, the nematode worm (Caenorhabditis elegans) has only been used as a model organism since the early 1960s. Why use the fly in research?FACTSThe fruit fly (Drosophila melanogaster) is one of the most well understood of all the model organisms.Why use yeast in research?FACTSBaker’s yeast, or Saccharomyces cerevisiae as it is also known, is among the best-studied experimental organisms.Inbreeding: from champion horses to life-saving miceSTORIESHumans have been breeding animals for millenia to bring out desirable characteristics. With the thoroughbred race horse there’s lots of money at stake but with research mice it’s the possibility of life-saving new treatments.

Logo for Kent Scientific

search There are no items in your cart.View Products ›888-572-8887

Mice vs Rats in Research: What’s the Difference?

Posted on 4/25/2019

Rodents, usually rats and mice, have been the most commonly used animals for biomedical research for more than a century for a number of reasons: they are readily available, easy to handle, and very similar to humans physiologically and genetically.

While there are similarities between mice and rats, there are several significant physiological and behavioral differences between the two that researchers need to consider when deciding which to use for a specific application.

Size: Adult rats weigh up to ten times more than adult mice, which makes them the preferred model for applications that involve surgical procedures, especially involving the brain and spinal cord. Surgery is generally easier in a larger animal, and causes less tissue damage. Imaging can also be more effective with rats because the larger size of the animal offers better resolution. The smaller size of mice does offer some advantages, including requiring lower drug dosages, which makes them more cost-effective in drug development research. Mice are also better models for optogenetics, since their smaller brain size makes it easier for light to reach deeper brain regions.

Handling: Rats are easier to handle and show less stress when being handled by humans. They can also be trained to hold still for some procedures, eliminating the need for anesthesia. Mice are more prone to stress from repeated handling and are more likely to need sedation for procedures.

Social behavior: Rats and mice are quite different in social cognition, which can be important in research for disorders including schizophrenia, attention deficit disorder (ADD), autism spectrum disorder (ASD), and mood disorders. Rats tend to enjoy being with other rats and are less territorial and less aggressive in social situations. Mice are generally more averse to interaction with other mice and show more aggressive behavior in social situations with other mice.

Addiction and impulsivity: The brains of mice and rats handle the delivery of serotonin and dopamine differently, which means they behave differently in terms of habit formation, impulsivity and other situations related to mood, addiction, and some psychiatric disorders. Rats show more compulsive and additive behavior, more impulsive behavior, and less self-restraint in waiting for a known reward. Mice are less impulsive and more able to control premature responses. Mice and rats have been shown to react very differently in their response to alcohol, nicotine, and MDMA.

Cognitive behavior: Cognition in rodents is often studied in relation to dementia, schizophrenia, and other psychiatric and neurological disorders in humans. Rats are superior at maze-learning, show a higher level of strategy, and show more stable performance in longer cognitive tests. Mice exhibit less strategy in their maze-learning, often needing substantially more training and practice to learn a maze, and experience more stress and anxiety while doing so.

Over the last two decades, researchers have increasingly used mice models more often than rats, primarily because of their superiority as genetic models. The first recombinant mouse model was identified in 1987, compared to 2010 for the first recombinant rat model. Now that gene-editing technologies are available for both mice and rats, researchers will have even more options in their quest to find the right rodent model for each research application.

‹ Back to Blog

Additional Questions?

Our product specialists are here to help with additional information on our products, grant proposals, orders and more.Call 888-572-8887

Email Signup

Facebook
LinkedIn
YouTube
Twitter

Kent Scientific Corporation
1116 Litchfield Street
Torrington, CT 06790, USAToll Free: (888) 572-8887
Local: (860) 626-1172
Fax: (860) 626-1179Web Solutions© 2006 – 2019 Kent Scientific Corporation. All rights reserved.

https://www.jax.org/why-the-mouse ´´Why do researchers work with mice?
Isn’t there a high rate of clinical failure for drugs tested in mice?
Computer models for biology are sophisticated and accurate. Why do we still need to test on mice when we can just use computers?
We are getting large amounts of medical data from human patients. Why do we still need to use mouse models to develop precise treatments for humans?
What’s more effective: human or animal studies?
Why are mice considered excellent models for humans?
What is the future of mouse-based research? Will researchers still be using them 10 or 15 years from today?
Why mouse genetics?
I read about research findings in mice that don’t work in humans. What is being done to improve the medical yield from mouse-based research?
How did the lab mouse come to be?
What is a mouse model?
How can research in mice lead to new ways to prevent and treat disease?
What does the mouse teach us that we can’t learn from yeast, worms, insects and fish?
What regulatory bodies govern what we can do with mice? With other animals? Why doesn’t the USDA cover mice and rats?
I’ve heard we can grow human organs in the lab, why do we need mice for medical research?
How can the public be assured that researchers are using mice in a responsible way?
Why do we need to test drugs on mice when we can test on human cells and tissues? ´´

Jackson LaboratoryWhy do researchers work with mice?SCROLL DOWNIsn’t there a high rate of clinical failure for drugs tested in mice?Why do we still need to test on mice when we can just use computers?View all questions

  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  

Why do researchers work with mice?

Search…

Follow us:

  •  
  •  
  •  
  •  
  •  
  •  
UAR - Engaging and Informing since 2009

Home > What is animal research? > 10 Facts > Mouse


What is animal research?

Mouse

 

1.       Mice are used in 74.7% of procedures in animal research

Numbers relate to procedures on animals in Great Britain in 2013:

  • Rats, mice and other rodents, all purpose-bred laboratory species 84.5%
  • Fish, amphibians, reptiles and birds 16.3%
  • Sheep, cows, pigs and other large mammals 1.5%
  • Dogs and cats, all bred for research, no strays or unwanted pets can be used 0.12%
  • Primates, mainly marmoset and macaque monkeys 0.07%

Chimpanzees, orang-utans and gorillas have not been used in the UK for over 20 years and their use is now banned.

http://www.understandinganimalresearch.org.uk/animals/types-animals/

2.       98% of human genes have a comparable gene in mice

The mouse makes an excellent model for human disease because the organization of their DNA and their gene expression is similar to humans, with ninety-eight percent of human genes having a comparable gene in the mouse. They have similar reproductive and nervous systems to humans, and suffer from many of the same diseases such as cancer, diabetes and even anxiety. Manipulating their genes can lead them to develop other diseases that do not naturally affect them, and as a result research on mice has helped the understanding of both human physiology and the causes of disease.

http://www.animalresearch.info/en/designing-research/research-animals/mouse/

3.       Inbred strains of mice have been a disease model, long before the mouse genome project and transgenics

Inbred strains of mice were used as disease models, long before the mouse genome project and transgenics. There are a large number of laboratory strains available, and their long breeding history means that mice of a single laboratory strain are isogenic. This is useful in experiments, as it reduces natural variation between subjects. Some inbred strains are used for their predisposition to certain mutations or genetic diseases, while others are used for their general health and resistance to mutations.

http://www.animalresearch.info/en/designing-research/research-animals/mouse/

4.       It has been possible to clone mice since 1998

http://www.animalresearch.info/en/designing-research/research-animals/mouse/

5.        Studies with mice were awarded 30 Nobel Prizes

http://www.animalresearch.info/en/medical-advances/nobel-prizes/

6.       Mice helped with the discovery of HIV and its treatment

Bone marrow-Liver-Thymus (BLT) mice have become an important part of research into new treatments and vaccines. With bone marrow, liver and thymus tissue transplanted from humans, the mice can be regarded as having a human immune system. This allows them to become infected by HIV when introduced vaginally and exhibit many of the hallmarks of the disease in humans.

http://www.animalresearch.info/en/medical-advances/diseases-research/aids-hiv/

7.       Mice are often involved in fundamental research – how genes work, how cells function and how the body integrates information

Mice are used in a vast range of experiments, many of which are classified as fundamental research, investigating the physiology of mammals.

http://www.animalresearch.info/en/designing-research/research-animals/mouse/

8.       Without mice, organ transplant, IVF, most vaccines, cancer cures, our understanding of the immune system, wouldn’t be the same

http://www.animalresearch.info/en/designing-research/research-animals/mouse-gm/

http://www.animalresearch.info/en/designing-research/research-animals/mouse-immunodeficient/

http://www.animalresearch.info/en/designing-research/research-animals/mouse/

9.       Mice don’t sweat or vomit

This is one reason why poison is so effective at dispatching these creatures. They can’t rid it from their body because they don’t sweat or vomit!

http://www.got-bugs.com/quality-pest-control-atlanta-cumming/things-go-bump-night-10-rodent-facts-you-probably-didn%E2%80%99t-know

10.   Mice can jump a foot in the air, are great climbers and swimmers

Mice are talented gymnasts. You may have seen a mouse jump a foot into the air. But jumping is not all they can do. They are also great climbers and swimmers. But the most fascinating thing about them is that they can squeeze through openings as small as the size of a dime.

http://509stateofmind.com/5-rodent-facts-you-probably-didnt-know/TweetShareShare

Last edited: 22 April 2015 15:41

Latest News

More articles

Subscribe to our Newsletter

Find out about our newsletter


© Copyright 2019
Understanding Animal Research

Our on-line surveys are
powered by SmartSurvey

NEWS & OPINIONMAGAZINESUBJECTSMULTIMEDIACAREERSSUBSCRIBE

Gut Microbe Linked to Nonalcoholic Fatty Liver Disease

Gut Microbe Linked to Nonalcoholic Fatty Liver DiseaseResearchers find strains of the bacterium Klebsiella pneumoniae that produce high levels of alcohol in 60 percent of patients with the condition.

  1. Home
  2. Subjects
  3. Mouse Research

Mouse Research

Image of the Day: Tamed Gut Bacteria

Image of the Day: Tamed Gut Bacteria

Nicoletta Lanese | Aug 8, 2019

Curbing the growth of harmful bacteria in mouse microbiomes reduces the animals’ incidence of inflammation-related colorectal cancer.

Mouse Genetics Shape the Gut Microbiome More than Their Environment

Mouse Genetics Shape the Gut Microbiome More than Their Environment

Nicoletta Lanese | Aug 6, 2019

Neither the maternal microbiome nor housing conditions appear to permanently alter which microbes remain in the animals.

Brain’s Fluid Drains via Lymphatic Vessels at the Base of the Skull

Brain’s Fluid Drains via Lymphatic Vessels at the Base of the Skull

Abby Olena | Jul 24, 2019

Detailed imaging of the rodent central nervous system reveals new information about the route cerebrospinal fluid takes to leave the brain.

Prominent Mouse Genetics Center Could be Shuttered

Prominent Mouse Genetics Center Could be Shuttered

Shawna Williams | Jun 21, 2019

Staff at the UK’s Harwell Institute were notified that a strategy board recommended halting its academic work, but a final decision is months away.

Image of the Day: Life and Death

Image of the Day: Life and Death

Carolyn Wilke | Mar 22, 2019

When hair follicle stem cells lose their protein-based death cue, they take on a new role helping to repair wounds in skin.

CAR T Cells Treat Lupus in Mice

CAR T Cells Treat Lupus in Mice

Abby Olena | Mar 6, 2019

T cells modified to target disease-contributing B cells improved survival in two mouse models of the autoimmune disease.

An Algorithm to Predict the Age of Your Lab Mice

An Algorithm to Predict the Age of Your Lab Mice

Anthony King | Feb 1, 2019

Researchers develop an app that can estimate the biological age of a rodent from its mug shot—and could give a boost to the science of human aging in the process.

Image of the Day: Mouse Tails

Image of the Day: Mouse Tails

Carolyn Wilke | Jan 24, 2019

Genetic mutations create lab mice with unusually long and short tails.

First Successful Gene Drive in Mammals

First Successful Gene Drive in Mammals

Abby Olena | Jan 23, 2019

Researchers use a CRISPR-Cas9 strategy to expand a desired trait from 50 percent of mouse pups to about 72 percent.Page 1 of 10123456789Trending

New “Prime Editing” Method Makes Only Single-Stranded DNA Cuts

New “Prime Editing” Method Makes Only Single-Stranded DNA CutsOrganoids Don’t Accurately Model Human Brain Development

Opinion: Boycotting Elsevier Is Not Enough

Opinion: Boycotting Elsevier Is Not Enough

Opinion: The Nature of Social Inequalities in Great Britain

Opinion: The Nature of Social Inequalities in Great BritainMultimedia

Image of the Day: Sea Cucumber Hormone Therapy

Image of the Day: Sea Cucumber Hormone Therapy

Image of the Day: Flight Styles

Image of the Day: Flight Styles

The Scientist's Current Issue's Magazine Cover

OCTOBER 2019

Brain Fog

Air Pollution May Cause Cognitive DeclineSUBSCRIBE TODAY

Air Pollution May Damage People’s Brains

Air Pollution May Damage People’s BrainsContaminants in the atmosphere appear to have harmful effects on neurodevelopment and cognitive function.

Is It Time to Rethink Parkinson’s Pathology?

Is It Time to Rethink Parkinson’s Pathology?New evidence points to a waste-clearing problem in patients’ cells, rather than the accumulation of protein tangles, as the root cause of the neurodegenerative disease.

Why Immune Cells Extrude Webs of DNA and Protein

Why Immune Cells Extrude Webs of DNA and ProteinExtracellular webs expelled by neutrophils trap invading pathogens, but these newly discovered structures also have ties to autoimmunity and cancer.Sponsored ContentLabQuizzesWebinarsVideosInfographicseBooksTechEdgeSponsored Interactive Crossword Puzzle

Mouse Models for Disease Research

Mouse Models for Disease ResearchGenetically modified mice have revolutionized the biological sciences, helping to uncover countless mechanisms of physiological and pathological function, as well as being instrumental for testing potential intervention possibilities. Understanding how mouse models work goes a long way in helping each scientist find a model that can help them answer their own research questions.

Advances in Immune Cell Profiling

Advances in Immune Cell ProfilingLearn more about the role of the immune system in cancer, multiplexing immune cell profiling, compiling immune cell phenotypes, and high-throughput cell profiling!Mining the Epigenome: Microarrays for DNA MethylationLearn about how microarray technology offers a high-throughput, cost-effective route to examining DNA methylation with this poster from The Scientist’s Creative Services Division and Brooks Life Sciences!MarketplaceSponsored Product Updates

Pii to Manufacture FDA-Approved Hormone Therapy Injection Drug Product

Pii to Manufacture FDA-Approved Hormone Therapy Injection Drug ProductPharmaceutics International, Inc. (Pii), a Contract Development and Manufacturing Organization (CDMO) headquartered in Hunt Valley, Maryland, is pleased to announce the  commercial supply of Fulvestrant Injection 250mg/5ml drug product, which was recently approved by the FDA.

Bio-Rad Joins EMBL's Corporate Partnership Programme

Bio-Rad Joins EMBL’s Corporate Partnership ProgrammeBio-Rad Laboratories, Inc. (NYSE: BIO and BIOb), a global leader of life science research and clinical diagnostic products, announces that it has joined the European Molecular Biology Laboratory (EMBL) Corporate Partnership Programme, to help expand EMBL’s portfolio of innovative scientific training courses designed for budding young scientists.

DNASTAR Appoints Inqaba Biotec as African Distributor

DNASTAR Appoints Inqaba Biotec as African DistributorDNASTAR® has appointed Inqaba Biotechnical Industries (Pty) Ltd (trading as inqaba biotecTM) as a distributor of its DNA, RNA and protein sequence analysis software in Africa, effective immediately.

The Role of Ultrapure Water for HPLC Analysis

The Role of Ultrapure Water for HPLC AnalysisDownload this application note to learn about the benefits of using ultrapure water as a mobile phase for HPLC!Stay Connected with

E-NEWSLETTER SIGN-UP

Subscribe to receive The Scientist Daily E-Newsletter in your inbox!

FACEBOOK PAGES

THE SCIENTISTTHE SCIENTIST CAREERSTHE GENOME SCIENTISTTHE ENVIROSCIENTISTTHE CELL SCIENTISTTHE MICRO SCIENTISTTHE CANCER SCIENTISTTHE NEUROSCIENTISTABOUT & CONTACTPRIVACY POLICYJOB LISTINGSSUBSCRIBEADVERTISENOW PART OF THE LABX MEDIA GROUP:LAB MANAGER MAGAZINE|LABX|LABWRENCH

© 1986–2019 THE SCIENTIST. ALL RIGHTS RESERVED.

Skip to main contentBecome a MemberLog InScienceMag.orgSearch

Advertisement

SHARE

A new way to modify DNA, “prime editor” couples two enzymes, Cas9 (blue) and reverse transcriptase (red), to a guide RNA (green) that takes the complex to a specific place on DNA’s double helix (yellow and purple) and also holds the code for an insertion of new DNA at that spot. PEYTON RANDOLPH

New ‘prime’ genome editor could surpass CRISPR

By Jon CohenOct. 21, 2019 , 11:00 AM

CRISPR, an extraordinarily powerful genome-editing tool invented in 2012, can still be clumsy. It sometimes changes genes it shouldn’t, and it edits by hacking through both strands of DNA’s double helix, leaving the cell to clean up the mess—shortcomings that limit its use in basic research and agriculture and pose safety risks in medicine. But a new entrant in the race to refine CRISPR promises to steer around some of its biggest faults. “It’s a huge step in the right direction,” chemist George Church, a CRISPR pioneer at Harvard University, says about the work, which appears online today in Nature.

This newfangled CRISPR, dubbed “prime editing,” could make it possible to insert or delete specific sequences at genome targets with less collateral damage. “Prime editors offer more targeting flexibility and greater editing precision,” says David Liu, a chemist at the Broad Institute in Cambridge, Massachusetts, whose lab led the new study and earlier invented a popular CRISPR refinement called base editing.

Liu, his postdoc Andrew Anzalone, and co-workers tested variations of their prime editors on several human and mouse cells, performing more than 175 different edits. As a proof of principle, they created and then corrected the mutations that cause sickle cell anemia and Tay-Sachs disease, DNA aberrations that previous genome-editing systems either could not fix or only did so inefficiently. The edits occurred in a high percentage of cells and caused relatively few off-target changes. In its paper, the team claims the technology “in principle can correct about 89% of known pathogenic human genetic variants.”

SIGN UP FOR OUR DAILY NEWSLETTER

Get more great content like this delivered right to you!

Most CRISPR systems rely on a molecular complex that couples a guide RNA—which homes in on a specific location in the genome—with an enzyme, Cas9, that cuts both strands of DNA. During the cell’s efforts to reconnect the DNA, its repair machinery can introduce or delete nucleotides. Researchers can take advantage of the botched repair to knock out genes that, say, cause a disease. They can also hijack the inefficient repair process to add DNA—even an entire gene.

But double-stranded breaks are “genome vandalism,” Church says. As the cell attempts to repair the break it introduces insertions and deletions willy-nilly, sometimes creating unwanted—and even dangerous—mutations.

Liu’s earlier handwork, base editing, does not cut the double-stranded DNA but instead uses the CRISPR targeting apparatus to shuttle an additional enzyme to a desired sequence, where it converts a single nucleotide into another. Many genetic traits and diseases are caused by a single nucleotide change, so base editing offers a powerful alternative for biotechnology and medicine. But the method has limitations, and it, too, often introduces off-target mutations.

Prime editing steers around shortcomings of both techniques by heavily modifying the Cas9 protein and the guide RNA. The altered Cas9 only “nicks” a single strand of the double helix, instead of cutting both. The new guide, called a pegRNA, contains an RNA template for a new DNA sequence, to be added to the genome at the target location. That requires a second protein, attached to Cas9: a reverse transcriptase enzyme, which can make a new DNA strand from the RNA template and insert it at the nicked site.

Liu, who has already formed a company around the new technology, Prime Medicine, stresses that to gain a place in the editing toolkit, it will have to prove robust and useful in many labs. Delivering the large construct of RNA and enzymes into living cells will also be difficult, and no one has yet shown it can work in an animal model.

Fyodor Urnov, scientific director at the Innovative Genomics Institute in Berkeley, California, reviewed the paper for Nature and says it brought “one of those ‘yay, science!’ kind of moments.” Prime editing “well may become the way that disease-causing mutations are repaired,” he says. But, he adds, it’s too soon to be sure. The technique “just showed up this year.”Posted in: 

doi:10.1126/science.aaz9297

Jon Cohen

Jon Cohen

Jon is a staff writer for Science.

More from News

Enjoy reading News from Science? Subscribe today. If you have already subscribed, log into your News account.

Got a tip?

How to contact the news team

Advertisement

Advertisement

ScienceInsider

More ScienceInsider

Sifter

More SifterRead the Latest Issue of Science

25 October 2019

Vol 366, Issue 6464

Magazine Cover

Table of Contents

Get Our E-Alerts

Receive emails from Science. See full listScience Table of ContentsScience Daily NewsWeekly News RoundupScience Editor’s ChoiceFirst Release NotificationScience Careers Job SeekerCountryCountry*AfghanistanAland IslandsAlbaniaAlgeriaAndorraAngolaAnguillaAntarcticaAntigua and BarbudaArgentinaArmeniaArubaAustraliaAustriaAzerbaijanBahamasBahrainBangladeshBarbadosBelarusBelgiumBelizeBeninBermudaBhutanBolivia, Plurinational State ofBonaire, Sint Eustatius and SabaBosnia and HerzegovinaBotswanaBouvet IslandBrazilBritish Indian Ocean TerritoryBrunei DarussalamBulgariaBurkina FasoBurundiCambodiaCameroonCanadaCape VerdeCayman IslandsCentral African RepublicChadChileChinaChristmas IslandCocos (Keeling) IslandsColombiaComorosCongoCongo, The Democratic Republic of theCook IslandsCosta RicaCote D’IvoireCroatiaCubaCuraçaoCyprusCzech RepublicDenmarkDjiboutiDominicaDominican RepublicEcuadorEgyptEl SalvadorEquatorial GuineaEritreaEstoniaEthiopiaFalkland Islands (Malvinas)Faroe IslandsFijiFinlandFranceFrench GuianaFrench PolynesiaFrench Southern TerritoriesGabonGambiaGeorgiaGermanyGhanaGibraltarGreeceGreenlandGrenadaGuadeloupeGuatemalaGuernseyGuineaGuinea-BissauGuyanaHaitiHeard Island and Mcdonald IslandsHoly See (Vatican City State)HondurasHong KongHungaryIcelandIndiaIndonesiaIran, Islamic Republic ofIraqIrelandIsle of ManIsraelItalyJamaicaJapanJerseyJordanKazakhstanKenyaKiribatiKorea, Democratic People’s Republic ofKorea, Republic ofKuwaitKyrgyzstanLao People’s Democratic RepublicLatviaLebanonLesothoLiberiaLibyan Arab JamahiriyaLiechtensteinLithuaniaLuxembourgMacaoMacedonia, The Former Yugoslav Republic ofMadagascarMalawiMalaysiaMaldivesMaliMaltaMartiniqueMauritaniaMauritiusMayotteMexicoMoldova, Republic ofMonacoMongoliaMontenegroMontserratMoroccoMozambiqueMyanmarNamibiaNauruNepalNetherlandsNew CaledoniaNew ZealandNicaraguaNigerNigeriaNiueNorfolk IslandNorwayOmanPakistanPalestinianPanamaPapua New GuineaParaguayPeruPhilippinesPitcairnPolandPortugalQatarReunionRomaniaRussian FederationRWANDASaint Barthélemy Saint Helena, Ascension and Tristan da CunhaSaint Kitts and NevisSaint LuciaSaint Martin (French part)Saint Pierre and MiquelonSaint Vincent and the GrenadinesSamoaSan MarinoSao Tome and PrincipeSaudi ArabiaSenegalSerbiaSeychellesSierra LeoneSingaporeSint Maarten (Dutch part)SlovakiaSloveniaSolomon IslandsSomaliaSouth AfricaSouth Georgia and the South Sandwich IslandsSouth SudanSpainSri LankaSudanSurinameSvalbard and Jan MayenSwazilandSwedenSwitzerlandSyrian Arab RepublicTaiwanTajikistanTanzania, United Republic ofThailandTimor-LesteTogoTokelauTongaTrinidad and TobagoTunisiaTurkeyTurkmenistanTurks and Caicos IslandsTuvaluUgandaUkraineUnited Arab EmiratesUnited KingdomUnited StatesUruguayUzbekistanVanuatuVenezuela, Bolivarian Republic ofVietnamVirgin Islands, BritishWallis and FutunaWestern SaharaYemenZambiaZimbabwe

Email I agree to receive emails from AAAS/Science
and Science advertisers, including information on 
products, services, and special offers which may
include but are not limited to news, career
information, & upcoming events.Sign up today

Required fields are indicated by an asterisk (*)

AAAS

© 2019 American Association for the Advancement of Science. All rights Reserved. AAAS is a partner of HINARIAGORAOARECHORUSCLOCKSSCrossRef and COUNTER.


Sign inGet started

A Guide To Machine Learning Interview Questions And Answers

Sahiti Kappagantula

Sahiti KappagantulaFollowAug 30 · 22 min read

Machine Learning Interview Questions — Edureka

Ever since machines started learning and reasoning without human intervention, we’ve managed to reach an endless peak of technical evolution. Needless to say, the world has changed since Artificial Intelligence, Machine Learning and Deep learning were introduced and will continue to do so until the end of time. In this Machine Learning Interview Questions blog, I have collected the most frequently asked questions by interviewers. These questions are collected after consulting with Machine Learning Certification Training Experts.

In this article on Machine Learning Interview Questions, I will be discussing the top Machine Learning related questions asked in your interviews. So, for your better understanding I have divided this blog into the following 3 sections:

  1. Machine Learning Core Interview Questions
  2. Machine Learning Using Python Interview Question
  3. Machine Learning Scenario based Interview Question

Machine Learning Core Interview Question

Q1. What are the different types of Machine Learning?

There are three ways in which machines learn:

  1. Supervised Learning
  2. Unsupervised Learning
  3. Reinforcement Learning

Supervised Learning:

Supervised learning is a method in which the machine learns using labeled data.

  • It is like learning under the guidance of a teacher
  • The training dataset is like a teacher which is used to train the machine
  • Model is trained on a pre-defined dataset before it starts making decisions when given new data

Unsupervised Learning:

Unsupervised learning is a method in which the machine is trained on unlabelled data or without any guidance

  • It is like learning without a teacher.
  • The model learns through observation & finds structures in data.
  • Model is given a dataset and is left to automatically find patterns and relationships in that dataset by creating clusters.

Reinforcement Learning:

Reinforcement learning involves an agent that interacts with its environment by producing actions & discovers errors or rewards.

  • It is like being stuck in an isolated island, where you must explore the environment and learn how to live and adapt to the living conditions on your own.
  • Model learns through the hit and trial method
  • It learns on the basis of reward or penalty given for every action it performs

Q2. How would you explain Machine Learning to a school-going kid?

  • Suppose your friend invites you to his party where you meet total strangers. Since you have no idea about them, you will mentally classify them on the basis of gender, age group, dressing, etc.
  • In this scenario, the strangers represent unlabeled data and the process of classifying unlabeled data points is nothing but unsupervised learning.
  • Since you didn’t use any prior knowledge about people and classified them on-the-go, this becomes an unsupervised learning problem.

Q3. How does Deep Learning differ from Machine Learning?

Q4. Explain Classification and Regression

Q5. What do you understand by selection bias?

  • It is a statistical error that causes a bias in the sampling portion of an experiment.
  • The error causes one sampling group to be selected more often than other groups included in the experiment.
  • Selection bias may produce an inaccurate conclusion if the selection bias is not identified

Q6. What do you understand by Precision and Recall?

Let me explain you this with an analogy:

  • Imagine that, your girlfriend gave you a birthday surprise every year for the last 10 years. One day, your girlfriend asks you: ‘Sweetie, do you remember all the birthday surprises from me?’
  • To stay on good terms with your girlfriend, you need to recall all the 10 events from your memory. Therefore, recall is the ratio of the number of events you can correctly recall, to the total number of events.
  • If you can recall all 10 events correctly, then, your recall ratio is 1.0 (100%) and if you can recall 7 events correctly, your recall ratio is 0.7 (70%)

However, you might be wrong in some answers.

  • For example, let’s assume that you took 15 guesses out of which 10 were correct and 5 were wrong. This means that you can recall all events but not so precisely
  • Therefore, precision is the ratio of a number of events you can correctly recall, to the total number of events you can recall (mix of correct and wrong recalls).
  • From the above example (10 real events, 15 answers: 10 correct, 5 wrong), you get 100% recall but your precision is only 66.67% (10 / 15)

Q7. Explain false negative, false positive, true negative and true positive with a simple example.

Let’s consider a scenario of a fire emergency:

  • True Positive: If the alarm goes on in case of a fire.
    Fire is positive and prediction made by the system is true.
  • False Positive: If the alarm goes on, and there is no fire.
    System predicted fire to be positive which is a wrong prediction, hence the prediction is false.
  • False Negative: If the alarm does not ring but there was a fire.
    System predicted fire to be negative which was false since there was fire.
  • True Negative: If the alarm does not ring and there was no fire.
    The fire is negative and this prediction was true.

Q8. What is the Confusion Matrix?

A confusion matrix or an error matrix is a table which is used for summarizing the performance of a classification algorithm.

Consider the above table where:

  • TN = True Negative
  • TP = True Positive
  • FN = False Negative
  • FP = False Positive

Q9. What is the difference between inductive and deductive learning?

  • Inductive learning is the process of using observations to draw conclusions
  • Deductive learning is the process of using conclusions to form observations

Q10. How is KNN different from K-means clustering?

Q11. What is ROC curve and what does it represent?

Receiver Operating Characteristic curve (or ROC curve) is a fundamental tool for diagnostic test evaluation and is a plot of the true positive rate (Sensitivity) against the false positive rate (Specificity) for the different possible cut-off points of a diagnostic test.

  • It shows the tradeoff between sensitivity and specificity (any increase in sensitivity will be accompanied by a decrease in specificity).
  • The closer the curve follows the left-hand border and then the top border of the ROC space, the more accurate the test.
  • The closer the curve comes to the 45-degree diagonal of the ROC space, the less accurate the test.
  • The slope of the tangent line at a cutpoint gives the likelihood ratio (LR) for that value of the test.
  • The area under the curve is a measure of test accuracy.

Q12. What’s the difference between Type I and Type II error?

Q13. Is it better to have too many false positives or too many false negatives? Explain.

It depends on the question as well as on the domain for which we are trying to solve the problem. If you’re using Machine Learning in the domain of medical testing, then a false negative is very risky, since the report will not show any health problem when a person is actually unwell. Similarly, if Machine Learning is used in spam detection, then a false positive is very risky because the algorithm may classify an important email as spam.

Q14. Which is more important to you — model accuracy or model performance?

Well, you must know that model accuracy is only a subset of model performance. The accuracy of the model and performance of the model are directly proportional and hence better the performance of the model, more accurate are the predictions.

Q15. What is the difference between Gini Impurity and Entropy in a Decision Tree?

  • Gini Impurity and Entropy are the metrics used for deciding how to split a Decision Tree.
  • Gini measurement is the probability of a random sample being classified correctly if you randomly pick a label according to the distribution in the branch.
  • Entropy is a measurement to calculate the lack of information. You calculate the Information Gain (difference in entropies) by making a split. This measure helps to reduce the uncertainty about the output label.

Q16. What is the difference between Entropy and Information Gain?

  • Entropy is an indicator of how messy your data is. It decreases as you reach closer to the leaf node.
  • The Information Gain is based on the decrease in entropy after a dataset is split on an attribute. It keeps on increasing as you reach closer to the leaf node.

Q17. What is Overfitting? And how do you ensure you’re not overfitting with a model?

Over-fitting occurs when a model studies the training data to such an extent that it negatively influences the performance of the model on new data.

This means that the disturbance in the training data is recorded and learned as concepts by the model. But the problem here is that these concepts do not apply to the testing data and negatively impact the model’s ability to classify the new data, hence reducing the accuracy on the testing data.

Three main methods to avoid overfitting:

  • Collect more data so that the model can be trained with varied samples.
  • Use ensembling methods, such as Random Forest. It is based on the idea of bagging, which is used to reduce the variation in the predictions by combining the result of multiple Decision trees on different samples of the data set.
  • Choose the right algorithm.

Q18.Explain Ensemble learning technique in Machine Learning.

Ensemble learning is a technique that is used to create multiple Machine Learning models, which are then combined to produce more accurate results. A general Machine Learning model is built by using the entire training data set. However, in Ensemble Learning the training data set is split into multiple subsets, wherein each subset is used to build a separate model. After the models are trained, they are then combined to predict an outcome in such a way that the variance in the output is reduced.

Q19. What is bagging and boosting in Machine Learning?

Q20. How would you screen for outliers and what should you do if you find one?

The following methods can be used to screen outliers:

  1. Boxplot: A box plot represents the distribution of the data and its variability. The box plot contains the upper and lower quartiles, so the box basically spans the Inter-Quartile Range (IQR). One of the main reasons why box plots are used is to detect outliers in the data. Since the box plot spans the IQR, it detects the data points that lie outside this range. These data points are nothing but outliers.
  2. Probabilistic and statistical models: Statistical models such as normal distribution and exponential distribution can be used to detect any variations in the distribution of data points. If any data point is found outside the distribution range, it is rendered as an outlier.
  3. Linear models: Linear models such as logistic regression can be trained to flag outliers. In this manner, the model picks up the next outlier it sees.
  4. Proximity-based models: An example of this kind of model is the K-means clustering model wherein, data points form multiple or ‘k’ number of clusters based on features such as similarity or distance. Since similar data points form clusters, the outliers also form their own cluster. In this way, proximity-based models can easily help detect outliers.

How do you handle these outliers?

  • If your data set is huge and rich then you can risk dropping the outliers.
  • However, if your data set is small then you can cap the outliers, by setting a threshold percentile. For example, the data points that are above the 95th percentile can be used to cap the outliers.
  • Lastly, based on the data exploration stage, you can narrow down some rules and impute the outliers based on those business rules.

Q21. What are collinearity and multicollinearity?

  • Collinearity occurs when two predictor variables (e.g., x1 and x2) in a multiple regression have some correlation.
  • Multicollinearity occurs when more than two predictor variables (e.g., x1, x2, and x3) are inter-correlated.

Q22. What do you understand by Eigenvectors and Eigenvalues?

  • Eigenvectors: Eigenvectors are those vectors whose direction remains unchanged even when a linear transformation is performed on them.
  • Eigenvalues: Eigenvalue is the scalar that is used for the transformation of an Eigenvector.

In the above example, 3 is an Eigenvalue, with the original vector in the multiplication problem being an eigenvector.

The Eigenvector of a square matrix A is a nonzero vector x such that for some number λ, we have the following:

Ax = λx,

where λ is an Eigenvalue
So, in our example, λ = 3 and X = [1 1 2]

Q23. What is A/B Testing?

  • A/B is Statistical hypothesis testing for a randomized experiment with two variables A and B. It is used to compare two models that use different predictor variables in order to check which variable fits best for a given sample of data.
  • Consider a scenario where you’ve created two models (using different predictor variables) that can be used to recommend products for an e-commerce platform.
  • A/B Testing can be used to compare these two models to check which one best recommends products to a customer.

Q24. What is Cluster Sampling?

  • It is a process of randomly selecting intact groups within a defined population, sharing similar characteristics.
  • Cluster Sample is a probability sample where each sampling unit is a collection or cluster of elements.
  • For example, if you’re clustering the total number of managers in a set of companies, in that case, managers (samples) will represent elements and companies will represent clusters.

Q25. Running a binary classification tree algorithm is quite easy. But do you know how the tree decides on which variable to split at the root node and its succeeding child nodes?

  • Measures such as, Gini Index and Entropy can be used to decide which variable is best fitted for splitting the Decision Tree at the root node.
  • We can calculate Gini as following:
    Calculate Gini for sub-nodes, using the formula — sum of square of probability for success and failure (p²+q²).
  • Calculate Gini for split using weighted Gini score of each node of that split
  • Entropy is the measure of impurity or randomness in the data, (for binary class):

Here p and q is the probability of success and failure respectively in that node.

  • Entropy is zero when a node is homogeneous and is maximum when both the classes are present in a node at 50% — 50%. To sum it up, the entropy must be as low as possible in order to decide whether or not a variable is suitable as the root node.

Machine Learning With Python Questions

This set of Machine Learning interview questions deal with Python-related Machine Learning questions.

Q1. Name a few libraries in Python used for Data Analysis and Scientific Computations.

Here is a list of Python libraries mainly used for Data Analysis:

  • NumPy
  • SciPy
  • Pandas
  • SciKit
  • Matplotlib
  • Seaborn
  • Bokeh

Q2. Which library would you prefer for plotting in Python language: Seaborn or Matplotlib or Bokeh?

It depends on the visualization you’re trying to achieve. Each of these libraries is used for a specific purpose:

  • Matplotlib: Used for basic plotting like bars, pies, lines, scatter plots, etc
  • Seaborn: Is built on top of Matplotlib and Pandas to ease data plotting. It is used for statistical visualizations like creating heatmaps or showing the distribution of your data
  • Bokeh: Used for interactive visualization. In case your data is too complex and you haven’t found any “message” in the data, then use Bokeh to create interactive visualizations that will allow your viewers to explore the data themselves

Q3. How are NumPy and SciPy related?

  • NumPy is part of SciPy.
  • NumPy defines arrays along with some basic numerical functions like indexing, sorting, reshaping, etc.
  • SciPy implements computations such as numerical integration, optimization and machine learning using NumPy’s functionality.

Q4. What is the main difference between a Pandas series and a single-column DataFrame in Python?

Q5. How can you handle duplicate values in a dataset for a variable in Python?

Consider the following Python code:

bill_data=pd.read_csv("datasetsTelecom Data AnalysisBill.csv")
bill_data.shape
#Identify duplicates records in the data
Dupes = bill_data.duplicated()
sum(dupes)
#Removing Duplicates
bill_data_uniq = bill_data.drop_duplicates()

Q6. Write a basic Machine Learning program to check the accuracy of a model, by importing any dataset using any classifier?

#importing dataset
import sklearn
from sklearn import datasets
iris = datasets.load_iris()
X = iris.data
Y = iris.target

#splitting the dataset
from sklearn.cross_validation import train_test_split
X_train, Y_train, X_test, Y_test = train_test_split(X,Y, test_size = 0.5)

#Selecting Classifier
my_classifier = tree.DecisionTreeClassifier()
My_classifier.fit(X_train, Y_train)
predictions = my_classifier(X_test)
#check accuracy
From sklear.metrics import accuracy_score
print accuracy_score(y_test, predictions)

Machine Learning Scenario Based Questions

This set of Machine Learning interview questions deal with scenario-based Machine Learning questions.

Q1. You are given a data set consisting of variables having more than 30% missing values? Let’s say, out of 50 variables, 8 variables have missing values higher than 30%. How will you deal with them?

  • Assign a unique category to the missing values, who knows the missing values might uncover some trend.
  • We can remove them blatantly.
  • Or, we can sensibly check their distribution with the target variable, and if found any pattern we’ll keep those missing values and assign them a new category while removing others.

Q2. Write an SQL query that makes recommendations using the pages that your friends liked. Assume you have two tables: a two-column table of users and their friends, and a two-column table of users and the pages they liked. It should not recommend pages you already like.

SELECT f.user_id, l.page_id
FROM friend f JOIN like l
ON f.friend_id = l.user_id
WHERE l.page_id NOT IN (SELECT page_id FROM like
WHERE user_id = f.user_id)

Q3. There’s a game where you are asked to roll two fair six-sided dice. If the sum of the values on the dice equals seven, then you win $21. However, you must pay $5 to play each time you roll both dice. Do you play this game? And in the follow-up: If he plays 6 times what is the probability of making money from this game?

  • The first condition states that if the sum of the values on the 2 dices is equal to 7, then you win $21. But for all the other cases you must pay $5.
  • First, let’s calculate the number of possible cases. Since we have two 6-sided dices, the total number of cases => 6*6 = 36.
  • Out of 36 cases, we must calculate the number of cases that produces a sum of 7 (in such a way that the sum of the values on the 2 dices is equal to 7)
  • Possible combinations that produce a sum of 7 is, (1,6), (2,5), (3,4), (4,3), (5,2) and (6,1). All these 6 combinations generate a sum of 7.
  • This means that out of 36 chances, only 6 will produce a sum of 7. On taking the ratio, we get: 6/36 = 1/6
  • So this suggests that we have a chance of winning $21, once in 6 games.
  • So to answer the question if a person plays 6 times, he will win one game of $21, whereas for the other 5 games he will have to pay $5 each, which is $25 for all five games. Therefore, he will face a loss because he wins $21 but ends up paying $25.

Q4. We have two options for serving ads within Newsfeed:
1 — out of every 25 stories, one will be an ad
2 — every story has a 4% chance of being an ad

For each option, what is the expected number of ads shown in 100 news stories?
If we go with option 2, what is the chance a user will be shown only a single ad in 100 stories? What about no ads at all?

  • The expected number of ads shown in 100 new stories for option 1 is equal to 4 (100/25 = 4).
  • Similarly, for option 2, the expected number of ads shown in 100 new stories is also equal to 4 (4/100 = 1/25 which suggests that one out of every 25 stories will be an ad, therefore in 100 new stories there will be 4 ads)
  • Therefore for each option, the total number of ads shown in 100 new stories is 4.
  • The second part of the question can be solved by using Binomial distribution. Binomial distribution takes three parameters:
  • The probability of success and failure, which in our case is 4%.
  • The total number of cases, which is 100 in our case.
  • The probability of the outcome, which is a chance that a user will be shown only a single ad in 100 stories
  • p(single ad) = (0.96)⁹⁹*(0.04)¹

(note: here 0.96 denotes the chance of not seeing an ad in 100 stories, 99 denotes the possibility of seeing only 1 ad, 0.04 is the probability of seeing an ad once in 100 stories )

  • In total, there are 100 positions for the ad. Therefore, 100 * p(single ad) = 7.03%

Q5. How would you predict who will renew their subscription next month? What data would you need to solve this? What analysis would you do? Would you build predictive models? If so, which algorithms?

  • Let’s assume that we’re trying to predict renewal rate for Netflix subscription. So our problem statement is to predict which users will renew their subscription plan for the next month.
  • Next, we must understand the data that is needed to solve this problem. In this case, we need to check the number of hours the channel is active for each household, the number of adults in the household, number of kids, which channels are streamed the most, how much time is spent on each channel, how much has the watch rate varied from last month, etc. Such data is needed to predict whether or not a person will continue the subscription for the upcoming month.
  • After collecting this data, it is important that you find patterns and correlations. For example, we know that if a household has kids, then they are more likely to subscribe. Similarly, by studying the watch rate of the previous month, you can predict whether a person is still interested in a subscription. Such trends must be studied.
  • The next step is analysis. For this kind of problem statement, you must use a classification algorithm that classifies customers into 2 groups:
  • Customers who are likely to subscribe next month
  • Customers who are not likely to subscribe next month
  • Would you build predictive models? Yes, in order to achieve this you must build a predictive model that classifies the customers into 2 classes like mentioned above.
  • Which algorithms to choose? You can choose classification algorithms such as Logistic Regression, Random Forest, Support Vector Machine, etc.
  • Once you’ve opted the right algorithm, you must perform model evaluation to calculate the efficiency of the algorithm. This is followed by deployment.

Q6. How do you map nicknames (Pete, Andy, Nick, Rob, etc) to real names?

  • This problem can be solved in n number of ways. Let’s assume that you’re given a data set containing 1000s of twitter interactions. You will begin by studying the relationship between two people by carefully analyzing the words used in the tweets.
  • This kind of problem statement can be solved by implementing Text Mining using Natural Language Processing techniques, wherein each word in a sentence is broken down and co-relations between various words are found.
  • NLP is actively used in understanding customer feedback, performing sentimental analysis on Twitter and Facebook. Thus, one of the ways to solve this problem is through Text Mining and Natural Language Processing techniques.

Q7. A jar has 1000 coins, of which 999 are fair and 1 is double headed. Pick a coin at random, and toss it 10 times. Given that you see 10 heads, what is the probability that the next toss of that coin is also a head?

  • There are two ways of choosing a coin. One is to pick a fair coin and the other is to pick the one with two heads.
  • Probability of selecting fair coin = 999/1000 = 0.999
    Probability of selecting unfair coin = 1/1000 = 0.001
  • Selecting 10 heads in a row = Selecting fair coin * Getting 10 heads + Selecting an unfair coin
  • P (A) = 0.999 * (1/2)¹⁰ = 0.999 * (1/1024) = 0.000976
    P (B) = 0.001 * 1 = 0.001
    P( A / A + B ) = 0.000976 / (0.000976 + 0.001) = 0.4939
    P( B / A + B ) = 0.001 / 0.001976 = 0.5061
  • Probability of selecting another head = P(A/A+B) * 0.5 + P(B/A+B) * 1 = 0.4939 * 0.5 + 0.5061 = 0.7531

Q8. Suppose you are given a data set which has missing values spread along 1 standard deviation from the median. What percentage of data would remain unaffected and Why?

Since the data is spread across the median, let’s assume it’s a normal distribution.
As you know, in a normal distribution, ~68% of the data lies in 1 standard deviation from mean (or mode, median), which leaves ~32% of the data unaffected. Therefore, ~32% of the data would remain unaffected by missing values.

Q9. You are given a cancer detection data set. Let’s suppose when you build a classification model you achieved an accuracy of 96%. Why shouldn’t you be happy with your model performance? What can you do about it?

You can do the following:

  • Add more data
  • Treat missing outlier values
  • Feature Engineering
  • Feature Selection
  • Multiple Algorithms
  • Algorithm Tuning
  • Ensemble Method
  • Cross-Validation

Q10. You are working on a time series data set. Your manager has asked you to build a high accuracy model. You start with the decision tree algorithm since you know it works fairly well on all kinds of data. Later, you tried a time series regression model and got higher accuracy than the decision tree model. Can this happen? Why?

  • Time series data is based on linearity while a decision tree algorithm is known to work best to detect non-linear interactions
  • Decision tree fails to provide robust predictions. Why?
  1. The reason is that it couldn’t map the linear relationship as good as a regression model did.
  2. We also know that a linear regression model can provide a robust prediction only if the data set satisfies its linearity assumptions.

Q11. Suppose you found that your model is suffering from low bias and high variance. Which algorithm you think could tackle this situation and Why?

Type 1: How to tackle high variance?

  • Low bias occurs when the model’s predicted values are near to actual values.
  • In this case, we can use the bagging algorithm (eg: Random Forest) to tackle high variance problem.
  • Bagging algorithm will divide the data set into its subsets with repeated randomized sampling.
  • Once divided, these samples can be used to generate a set of models using a single learning algorithm. Later, the model predictions are combined using voting (classification) or averaging (regression).

Type 2: How to tackle high variance?

  • Lower the model complexity by using regularization technique, where higher model coefficients get penalized.
  • You can also use top n features from variable importance chart. It might be possible that with all the variable in the data set, the algorithm is facing difficulty in finding the meaningful signal.

Q12. You are given a data set. The data set contains many variables, some of which are highly correlated and you know about it. Your manager has asked you to run PCA. Would you remove correlated variables first? Why?

Possibly, you might get tempted to say no, but that would be incorrect.
Discarding correlated variables will have a substantial effect on PCA because, in the presence of correlated variables, the variance explained by a particular component gets inflated.

Q13. You are asked to build a multiple regression model but your model R² isn’t as good as you wanted. For improvement, you remove the intercept term now your model R² becomes 0.8 from 0.3. Is it possible? How?

Yes, it is possible.

  • The intercept term refers to model prediction without any independent variable or in other words, mean prediction
    R² = 1 — ∑(Y — Y´)²/∑(Y — Ymean)² where Y´ is the predicted value.
  • In the presence of the intercept term, R² value will evaluate your model with respect to the mean model.
  • In the absence of the intercept term (Ymean), the model can make no such evaluation,
  • With large denominator,
    Value of ∑(Y — Y´)²/∑(Y)² equation becomes smaller than actual, thereby resulting in a higher value of R².

Q14. You’re asked to build a random forest model with 10000 trees. During its training, you got training error as 0.00. But, on testing the validation error was 34.23. What is going on? Haven’t you trained your model perfectly?

  • The model is overfitting the data.
  • Training error of 0.00 means that the classifier has mimicked the training data patterns to an extent.
  • But when this classifier runs on the unseen sample, it was not able to find those patterns and returned the predictions with more number of errors.
  • In Random Forest, it usually happens when we use a larger number of trees than necessary. Hence, to avoid such situations, we should tune the number of trees using cross-validation.

Q15. ‘People who bought this also bought…’ recommendations seen on Amazon is based on which algorithm?

E-commerce websites like Amazon make use of Machine Learning to recommend products to their customers. The basic idea of this kind of recommendation comes from collaborative filtering. Collaborative filtering is the process of comparing users with similar shopping behaviors in order to recommend products to a new user with similar shopping behavior.

To better understand this, let’s look at an example. Let’s say a user A who is a sports enthusiast bought, pizza, pasta, and a coke. Now a couple of weeks later, another user B who rides a bicycle buys pizza and pasta. He does not buy the coke, but Amazon recommends a bottle of coke to user B since his shopping behaviors and his lifestyle is quite similar to user A. This is how collaborative filtering works.

With this, we come to an end of this blog. I hope these Machine Learning Interview Questions will help you ace your Machine Learning Interview. If you wish to check out more articles on the market’s most trending technologies like Python, DevOps, Ethical Hacking, then you can refer to Edureka’s official site.

Do look out for other articles in this series which will explain the various other aspects of Data Science.

1. Data Science Tutorial

2. Math And Statistics For Data Science

3. Machine Learning in R

4. Machine Learning Algorithms

5. Linear Regression In R

6. Logistic Regression in R

7. Classification Algorithms

8. Decision Tree in R

9. Introduction To Machine Learning

10. Naive Bayes in R

11. Statistics and Probability

12. Random Forest in R

13. Top 10 Myths Regarding Data Scientists Roles

14. Top Data Science Projects

15. Data Analyst vs Data Engineer vs Data Scientist

16. Types Of Artificial Intelligence

17. R vs Python

18. Artificial Intelligence vs Machine Learning vs Deep Learning


Originally published at https://www.edureka.co.Edureka

There are many e-learning platforms on the internet & then there’s us. We are not the biggest, but we are the fastest growing. We have the highest course completion rate in the industry. We provide live, instructor-led online programs in trending tech with 24×7 lifetime support.

Follow

17

 

 

17 claps

 

Sahiti Kappagantula

WRITTEN BY

Sahiti Kappagantula

Follow

A Data Science and Robotic Process Automation Enthusiast. Technical Writer.

Edureka

Edureka

Follow

There are many e-learning platforms on the internet & then there’s us. We are not the biggest, but we are the fastest growing. We have the highest course completion rate in the industry. We provide live, instructor-led online programs in trending tech with 24×7 lifetime support.

See responses (1)

More From Medium

More from Edureka

CI CD Pipeline: Learn How to Setup a CI CD Pipeline From Scratch

Saurabh Kulshrestha

Saurabh Kulshrestha inEdurekaAug 9, 2018 · 10 min read

1.8K

 

More from Edureka

Trees in Java — How to Implement a Binary Tree?

Swatee Chand

Swatee Chand in EdurekaSep 3 · 6 min read

58

 

More from Edureka

Java OOP Cheat Sheet — A Quick Guide to Object-Oriented Programming in Java

Swatee Chand

Swatee Chand in EdurekaAug 14 · 7 min read

134

Discover MediumWelcome to a place where words matter. On Medium, smart voices and original ideas take center stage – with no ads in sight. WatchMake Medium yoursFollow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. ExploreBecome a memberGet unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. UpgradeAboutHelpLegal

 WATCH NOW: FOX40 NEWS AT 5 FOX40 TV SCHEDULEAUTOS SEARCH  CONTACT USFOX40

SpaceX reveals early users of satellite-based high-speed internet

POSTED 8:25 AM, OCTOBER 26, 2019, BY CNN WIREUPDATED AT 08:24AM, OCTOBER 26, 2019

(CNN) — SpaceX is on a mission to beam cheap, high-speed internet to consumers all over the globe. And this week the company revealed a few earthly locations that are already linked to the network, including CEO Elon Musk’s house and the cockpits of a few Air Force jets.

It’s part of early testing for the 60 broadband-beaming satellites and two demo devices that SpaceX has already launched into orbit.

Eventually, the company wants to operate thousands of satellites that will circle the planet at about 300 to 700 miles overhead. The project is called Starlink, and if it’s successful it could forever alter the landscape of the telecom industry.

It may also bring in tens of billions of dollars for SpaceX each year if Starlink can compete with existing internet providers and help bring more people online. About half of the world’s population doesn’t have access to the internet, studies show.

A batch of 60 satellites launched atop one of SpaceX’s Falcon 9 rockets in May, and the company has laid out plans to rapidly build up the constellation. SpaceX plans as many as 24 dedicated Starlink launches — each with about 60 satellites — next year. Public filings show SpaceX wants to launch service in the southern United States next year.

In the meantime, the network is in testing mode.

Musk posted on Twitter Tuesday that said he was “sending this tweet through space via Starlink satellite.”

Starlink is a multibillion-dollar bet for SpaceX. Companies have tried and failed to deliver space-based internet before, but a new crop of ventures including SpaceX, Amazon and OneWeb is after it again.

SpaceX president and chief operating officer Gwynne Shotwell told reporters at a space conference in Washington DC that Musk already installed a Starlink user terminal at his home to test out the service.

She also revealed that the Air Force is also testing Starlink service on a C-12, an aircraft type used for passenger and cargo transport, and the military is working to add the service to other planes. So far, the service has been “one hundred times faster” than previous connections, Shotwell told reporters.

Shotwell went over other details about Starlink in an interview this week with reporters. Her quotes have been edited for length and clarity.

When’s the next launch?

The next launch that we have is a dedicated Starlink launch in mid-November, and those satellites will fly on a Falcon 9 booster that’s been launched three times before.

The 60 satellites that we already flew are capable of operations, but the next version will have upgraded technology. By late next year, we’ll be flying satellite with lasers that allow them to talk to each other in space and share data, which ensures customers will never lose service.

How will SpaceX compete with exiting internet providers in the US?

Is anybody paying less than 80 bucks a month for crappy service? Nope. That’s why we’re gonna be successful.

What will consumers need to connect to the Starlink network?

Consumers, hopefully, are going to receive a box with a user terminal and a cord, and it’ll sit either out of window on a roof or out on a kind of on a pole in your yard.

We’ve got a prototype, but we still have a lot of work to do. Hopefully, we’ll start to roll out service mid-next year.

Will SpaceX work with an existing telecom company?

In countries where we can, we are likely to go directly to consumers. We’ll have the full team of salespeople and tech support. Though, the better engineering that we do on the user terminal, the less service people we will need.

Is SpaceX concerned about getting permission to operate the service in other countries?

Right now, we’re focused on the United States and Canada. But there are a number of large organizations and governments that are quite interested in this capability.

It’s important for us to get the network up quickly.

What could hold up the Starlink launches -— satellite or rocket production?

I don’t think satellite production is going to be the holdup. It probably will be manufacturing the second stages of the rockets, or the rocket’s nose cone, or fairing, which protects the satellites during launch.

We’re working to reuse a fairing for the first time. And we’d like to fly Starlink missions exclusively with resued fairings.

Does SpaceX have an edge over Amazon and OneWeb?

I think there’s room for competition — perhaps one other space-based internet provider.

The existing telecom operators are trying to improve as well, but it’s hard for them to reach rural populations. Starlink by definition will cover all parts of the globe.

Note: At an investing conference in New York on Friday, Shotwell added this about OneWeb: “Our competitors are largely these new entrants to the market. OneWeb? We are 17 times better per bit. If you’re thinking about investing in OneWeb, I would recommend strongly against it”

Will SpaceX survive if Starlink fails?

We make money on Falcon rockets and Dragon spacecraft. Starlink is additive to our business.Play Video

TRADEMARK AND COPYRIGHT 2019 CABLE NEWS NETWORK, INC., A TIME WARNER COMPANY. ALL RIGHTS RESERVED.


FILED IN: BUSINESSNEWS SUGGEST A CORRECTION

by TaboolaSponsored LinksYou May LikeAt 80, Al Pacino Lives Modest Life With His PartnerCash RoadsterTop 30 Most Beautiful Women in the WorldHealthy Woman MagNo Chemicals, No Refills: This is the New Anti Mosquito Device You NeedMoskinatorWearing These Things On Airplane Is A Big MistakeTop Journey MagWhat Happens To Your Body If You Eat Bananas Every DayNutrition Expert17 Warning Signs Of Diabetes: What You Need To KnowHealth & Human ResearchPromoted LinksFROM AROUND THE WEBMORE FROM USGeography Facts That Will Blow Your Mind (Far & Wide)Richest Countries in the World, Ranked (Work + Money)These Hall of Famers Don’t Belong in Cooperstown (Stadium Talk)At 52, Salma Hayek Is Unrecognizable Today (Medical Matters)Most Popular Dog Breeds in the U.S., Ranked (FamilyMinded)Former ‘To Catch a Predator’ Host Chris Hansen Arrested in ConnecticutElon Musk’s ‘Not a Flamethrower’ Devices Show Up on eBay for ThousandsSpaceX Launches 1st Recycled Supply Ship26 Teenage Girls Found Dead at SeaFour Charged After Allegedly Raping 9-Year-Old Girl While Mom Smoked Meth in Garageby TaboolaFinal Quarter Friday Night Fan FavoriteIt’s that time again! Which high school football game should FOX40 and 106.5 THE END cover as the Friday Night Fan Favorite?

POPULAR
LATEST NEWS

ONLINE PUBLIC FILE • TERMS OF SERVICE • PRIVACY POLICY • 4655 FRUITRIDGE ROAD SACRAMENTO, CA 95820 • COPYRIGHT © 2019, KTXL • NEXSTAR BROADCASTING, INC. | ALL RIGHTS RESERVED • POWERED BY WORDPRESS.COM VIP

  •  
  •  

Pular para a barra de ferramentas

Sair

 

Edition

New gene editing technology could correct 89% of genetic defects

 

By Jessie Yeung, CNN

Updated 0836 GMT (1636 HKT) October 22, 2019

CRISPR Cas9 explainer natpkg_00010203
  •  

Play VideoWhat is CRISPR Cas9 gene editing? 01:50

(CNN)Scientists have developed a new gene-editing technology that could potentially correct up to 89% of genetic defects, including those that cause diseases like sickle cell anemia.The new technique is called “prime editing,” and was developed by researchers from the Broad Institute of MIT and Harvard, who published their findings Monday in the journal Nature.Prime editing builds on powerful CRISPR gene editing, but is more precise and versatile — it “directly writes new genetic information into a specified DNA site,” according to the paper.In the traditional CRISPR-Cas9 approach, Cas9, a type of modified protein, acts like a pair of scissors that can snip parts of DNA strands. It can target genes in a specific location — for instance, to disrupt a mutation.About two-thirds of known human genetic variants associated with diseases are single point gene mutations, so gene editing has the potential to correct or reproduce such mutations.Scientists edit gene for blood disease in human embryosPrime editing combines the CRISPR-Cas9 method with a different protein that can generate new DNA. The tool nicks the DNA strand, then transfers an edited sequence to the target DNA — allowing researchers to smoothly insert and delete parts of human cells.The technique allows researchers to search and replace entire sections of DNA strands, all without disruptive breaks or donor DNA. With this method, researchers say they hope to accurately and efficiently correct up to 89% of known disease-causing genetic variations.”With prime editing, we can now directly correct the sickle-cell anemia mutation back to the normal sequence and remove the four extra DNA bases that cause Tay Sachs disease, without cutting DNA entirely or needing DNA templates,” said David Liu, one of the authors of the study, in a Broad Institute press release.The scientist, the twins and the experiment that geneticists say went too far“The versatility of prime editing quickly became apparent as we developed this technology,” said Andrew Anzalone, another author in the study, in the press release. “The fact that we could directly copy new genetic information into a target site was a revelation. We were really excited.”The team of researchers will now continue working to hone the technique, trying to maximize its efficiency in various cell types and exploring any potential effects on the cells. They will also continue testing on different models of diseases to ultimately “provide a potential path for human therapeutic applications,” according to the press release.Gene editing is still a relatively young and rapidly expanding field of study — CRISPR-Cas9 is based on a decade-old discovery, but was only used on humans for the first time in 2016. Then in 2017, the Broad Institute developed a new technique called base editing, which can make changes to a targeted DNA site without cutting the DNA.Proposal for global moratorium on editing of inherited DNA is met with criticismResearchers at the Broad Institute and elsewhere hope CRISPR could one day target a wide range of “bad” genes — potentially helping humans avoid obesity, Alzheimer’s disease, genetic forms of deafness, and more.However, as the technology has advanced, doctors, scientists, and bioethicists have also raised ethical questions. Some fear it could open the door to human embryos being manipulated for nontherapeutic reasons, or that it could create unintended mutations and new diseases.Just earlier this year in March, a group of researchers, including the scientist who pioneered and patented CRISPR technology, called for a global moratorium on human germline editing — changes made to inherited DNA that can be passed on to the next generation.They listed ethical concerns, and pointed to Chinese scientist He Jiankui, who claimed to have made gene edits when creating two AIDS-resistant babies last year. He’s work, which could have unforeseen consequences, has been internationally condemned and called “abominable in nature” by Chinese authorities.

  •  

Search

FOLLOW CNN

  •  
  •  
  •  

© 2019 Cable News Network.Turner Broadcasting System, Inc.All Rights Reserved.CNN Sans ™ & © 2016 Cable News Network.Barack Obama and Donald Trump wanted to watch when two of the world’s biggest terrorists were killed. The similarities end there

rodent research models

eTACONIC®
ORDERCONTACT

 |  | 

CRISPR GENE EDITING

Request a free
MODEL GENERATION CONSULTATION

 GO

CRISPR gene editing speeds timelines, reduces costs, and improves efficiencies in custom model generation.

  • Dramatically reduce timelines compared to traditional gene targeting methods.
  • Efficiently retarget existing animal models and repositories.
  • Generate custom knockouts and point mutations at significant cost reduction.
CRISPR How It Works

Taconic Biosciences uses the latest CRISPR/Cas9 gene editing technology to generate genetically-engineered animal models with a 100% success rate. Taconic holds CRISPR licenses from both UC Berkeley and the Broad Institute.Contact us to talk with a Taconic scientist about applying CRISPR technologies to your animal model design program.

Generating CRISPR Mice and Rats

Taconic offers a mature, successful program for the design and generation of CRISPR-edited animal models.

Your CRISPR founders, fully characterized via genotyping and sequencing data, can be seamlessly transitioned to Taconic’s rapid colony expansion and colony management programs.

 

  • Selection of target site, assays, and oligonucleotides, as well as predicted off-target sites.
  • Preparation of sgRNA and Cas9 mRNA.
  • Injection of sgRNA and Cas9 mRNA into mouse C57BL/6NTac zygotes, to generate F0 mutated candidate animals.
  • Genotyping and target-site sequence analysis in mutated candidate F0 animals.
  • Optional steps:
    • Germline breeding of F0 founder animal (optional)
    • Sequence analysis of germline transmitted F1 animals (optional)
    • Off-target analysis in germline transmitted F1 animals (optional)
Related White Paper

Download the Taconic Biosciences’ White Paper:

The CIEA NOG mouse

Download the Taconic Biosciences’ Poster:

At Taconic Biosciences, we use both an in vivo strategy utilizing one-cell embryos and a complementary in vitro strategy utilizing embryonic stem (ES) cells to generate both mouse and rat models with CRISPR/Cas9. We provide a broad data-set to illustrate our experiences using these two strategies within our production pipeline.

Genetically Engineered Model Publications Database

Search Taconic’s GEMs Design database by application or model type:

GEMs Design inquiry

Talk to a Taconic Biosciences scientist about designing a custom model:

DESIGN YOUR MODEL

pagedata icon

ASK US ABOUT OUR CRISPR SERVICES

Schedule a consultation:Select your prefixDr.Mr.Mrs.Ms.Prof.Select Your Job RoleAcademic/Government ResearcherBiotech Group DirectorBiotech ResearcherCRO DirectorCRO Research AssociateFor Profit ExecutiveFor Profit Laboratory Animal ResourcesNonprofit ExecutiveNonprofit Laboratory Animal ResourcesNot-For-Profit ResearcherPharma DirectorPharma ResearcherProcurement OfficerVeterinarianSelect Main Research AreaADMETAlzheimer’s DiseaseAnimal WelfareCardiovascular DiseaseColony Management SolutionsCryopreservationGeneticsInfectious DiseaseInflammationImmunologyMetabolic DiseaseModel Generation SolutionsMicrobiome & Germ-FreeNeuroscienceOncology & Immuno-OncologyParkinson’s DiseaseQualityOtherSelect Your CountryAfghanistanAlbaniaAlgeriaAndorraAngolaAntigua & DepsArgentinaArmeniaAustraliaAustriaAzerbaijanBahamasBahrainBangladeshBarbadosBelarusBelgiumBelizeBeninBhutanBoliviaBosnia HerzegovinaBotswanaBrazilBruneiBulgariaBurkinaBurundiCambodiaCameroonCanadaCape VerdeCentral African RepChadChileChinaColombiaComorosCongoCongo {Democratic Rep}Costa RicaCroatiaCubaCyprusCzech RepublicDenmarkDjiboutiDominicaDominican RepublicEast TimorEcuadorEgyptEl SalvadorEquatorial GuineaEritreaEstoniaEthiopiaFijiFinlandFranceGabonGambiaGeorgiaGermanyGhanaGreeceGrenadaGuatemalaGuineaGuinea-BissauGuyanaHaitiHondurasHungaryIcelandIndiaIndonesiaIranIraqIreland {Republic}IsraelItalyIvory CoastJamaicaJapanJordanKazakhstanKenyaKiribatiKorea NorthKorea SouthKosovoKuwaitKyrgyzstanLaosLatviaLebanonLesothoLiberiaLibyaLiechtensteinLithuaniaLuxembourgMacedoniaMadagascarMalawiMalaysiaMaldivesMaliMaltaMarshall IslandsMauritaniaMauritiusMexicoMicronesiaMoldovaMonacoMongoliaMontenegroMoroccoMozambiqueMyanmar, (Burma)NamibiaNauruNepalNetherlandsNew ZealandNicaraguaNigerNigeriaNorwayOmanPakistanPalauPanamaPapua New GuineaParaguayPeruPhilippinesPolandPortugalQatarRomaniaRussian FederationRwandaSt Kitts & NevisSt LuciaSaint Vincent & the GrenadinesSamoaSan MarinoSao Tome & PrincipeSaudi ArabiaSenegalSerbiaSeychellesSierra LeoneSingaporeSlovakiaSloveniaSolomon IslandsSomaliaSouth AfricaSouth SudanSpainSri LankaSudanSurinameSwazilandSwedenSwitzerlandSyriaTaiwanTajikistanTanzaniaThailandTogoTongaTrinidad & TobagoTunisiaTurkeyTurkmenistanTuvaluUgandaUkraineUnited Arab EmiratesUnited StatesUnited KingdomUruguayUzbekistanVanuatuVatican CityVenezuelaVietnamYemenZambiaZimbabwe

WHAT OUR CUSTOMERS SAY:

“We are very pleased to add Taconic to our list of licensees as it furthers our goal of making this technology more broadly accessible and Taconic is certainly a globally recognized leader in the animal model field.”
Eric Rhodes,
CEO of ERS
Genomics

© 2019 Taconic Biosciences, Inc. All rights reserved. Sitemap

  1. Home
  2. Blog
  3. Data Science
  4. Top Machine Learning Interview…

Data Science with Python (19 Blogs)Become a Certified Professional 

Introduction to Python
Python Installation
Python Fundamentals
Python OOPs
Python Libraries
Web Scraping
Django
Python Programs
Career Oppurtunities
Interview Questions

Data Science

Topics Covered

  • Business Analytics with R (30 Blogs)
  • Data Science (39 Blogs)
  • Mastering Python (66 Blogs)
  • Decision Tree Modeling Using R (1 Blogs)

SEE MORE 

Top Machine Learning Interview Questions You Must Prepare In 2019

 Last updated on Sep 03,201919.7K Views

Zulaikha Lateef

Zulaikha LateefZulaikha is a tech enthusiast working as a Research Analyst at Edureka.

myMock Interview Service for Real Tech JobsMock interview in latest tech domains i.e JAVA, AI, DEVOPS,etcGet interviewed by leading tech expertsReal time assessment report and video recordingTRY OUT MOCK INTERVIEW

Ever since machines started learning and reasoning without human intervention, we’ve managed to reach an endless peak of technical evolution. Needless to say, the world has changed since Artificial IntelligenceMachine Learning and Deep learning were introduced and will continue to do so until the end of time. In this Machine Learning Interview Questions blog, I have collected the most frequently asked questions by interviewers. These questions are collected after consulting with Machine Learning Certification Training Experts.

In case you have attended any Machine Learning interview in the recent past, do paste those interview questions in the comments section and we’ll answer them at the earliest. You can also comment below if you have any questions in your mind, which you might face in your Machine Learning interview.

You may go through this recording of Machine Learning Interview Questions and Answers where our instructor has explained the topics in a detailed manner with examples that will help you to understand this concept better.

Machine Learning Interview Questions and Answers | Edureka

In this blog on Machine Learning Interview Questions, I will be discussing the top Machine Learning related questions asked in your interviews. So, for your better understanding I have divided this blog into the following 3 sections:

  1. Machine Learning Core Interview Questions
  2. Machine Learning Using Python Interview Question
  3. Machine Learning Scenario based Interview Question

Machine Learning Core Interview Question

Q1. What are the different types of Machine Learning?

Types of Machine Learning - Machine Learning Interview Questions - Edureka

Types of Machine Learning – Machine Learning Interview Questions – Edureka

There are three ways in which machines learn:

  1. Supervised Learning
  2. Unsupervised Learning
  3. Reinforcement Learning

Supervised Learning:
Supervised learning is a method in which the machine learns using labeled data. 

  • It is like learning under the guidance of a teacher
  • Training dataset is like a teacher which is used to train the machine
  • Model is trained on a pre-defined dataset before it starts making decisions when given new data

Unsupervised Learning:
 Unsupervised learning is a method in which the machine is trained on unlabelled data or without any guidance 

  • It is like learning without a teacher.
  • Model learns through observation & finds structures in data.
  • Model is given a dataset and is left to automatically find patterns and relationships in that dataset by creating clusters.

Reinforcement Learning:
 Reinforcement learning involves an agent that interacts with its environment by producing actions & discovers errors or rewards. 

  • It is like being stuck in an isolated island, where you must explore the environment and learn how to live and adapt to the living conditions on your own.
  • Model learns through the hit and trial method
  • It learns on the basis of reward or penalty given for every action it performs

Q2. How would you explain Machine Learning to a school-going kid?

  • Suppose your friend invites you to his party where you meet total strangers. Since you have no idea about them, you will mentally classify them on the basis of gender, age group, dressing, etc.
  • In this scenario, the strangers represent unlabeled data and the process of classifying unlabeled data points is nothing but unsupervised learning.
  • Since you didn’t use any prior knowledge about people and classified them on-the-go, this becomes an unsupervised learning problem.

Q3. How does Deep Learning differ from Machine Learning?

Deep LearningMachine Learning
Deep Learning is a form of machine learning that is inspired by the structure of the human brain and is particularly effective in feature detection.Machine Learning is all about algorithms that parse data, learn from that data, and then apply what they’ve learned to make informed decisions.

Deep Learning vs Machine Learning – Machine Learning Interview Questions – Edureka

Q4. Explain Classification and Regression

Classification vs Regression - Machine Learning Interview Questions - Edureka

Classification vs Regression – Machine Learning Interview Questions – Edureka

Q5. What do you understand by selection bias?

  • It is a statistical error that causes a bias in the sampling portion of an experiment.
  • The error causes one sampling group to be selected more often than other groups included in the experiment.
  • Selection bias may produce an inaccurate conclusion if the selection bias is not identified.

Q6. What do you understand by Precision and Recall?

Let me explain you this with an analogy:

  • Imagine that, your girlfriend gave you a birthday surprise every year for the last 10 years. One day, your girlfriend asks you: ‘Sweetie, do you remember all the birthday surprises from me?’
  • To stay on good terms with your girlfriend, you need to recall all the 10 events from your memory. Therefore, recall is the ratio of the number of events you can correctly recall, to the total number of events.
  • If you can recall all 10 events correctly, then, your recall ratio is 1.0 (100%) and if you can recall 7 events correctly, your recall ratio is 0.7 (70%)

However, you might be wrong in some answers.

  • For example, let’s assume that you took 15 guesses out of which 10 were correct and 5 were wrong. This means that you can recall all events but not so precisely
  • Therefore, precision is the ratio of a number of events you can correctly recall, to the total number of events you can recall (mix of correct and wrong recalls).
  • From the above example (10 real events, 15 answers: 10 correct, 5 wrong), you get 100% recall but your precision is only 66.67% (10 / 15)

Q7. Explain false negative, false positive, true negative and true positive with a simple example.

Let’s consider a scenario of a fire emergency:

  • True Positive: If the alarm goes on in case of a fire.
    Fire is positive and prediction made by the system is true.
  • False Positive: If the alarm goes on, and there is no fire.
    System predicted fire to be positive which is a wrong prediction, hence the prediction is false.
  • False Negative: If the alarm does not ring but there was a fire.
    System predicted fire to be negative which was false since there was fire.
  • True Negative: If the alarm does not ring and there was no fire.
    The fire is negative and this prediction was true.

Q8. What is a Confusion Matrix? 

A confusion matrix or an error matrix is a table which is used for summarizing the performance of a classification algorithm. 

Confusion Matrix - Machine Learning Interview Questions - Edureka

Confusion Matrix – Machine Learning Interview Questions – Edureka

Consider the above table where:

  • TN = True Negative
  • TP = True Positive
  • FN = False Negative
  • FP = False Positive

Q9. What is the difference between inductive and deductive learning?

  • Inductive learning is the process of using observations to draw conclusions 
  • Deductive learning is the process of using conclusions to form observations 
Inductive & Deductive learning - Machine Learning Interview Questions - Edureka

Inductive vs Deductive learning – Machine Learning Interview Questions – Edureka

Q10. How is KNN different from K-means clustering?

Kmeans vs KNN - Machine Learning Interview Questions - Edureka

K-means vs KNN – Machine Learning Interview Questions – EdurekaMachine Learning Certification Training using PythonInstructor-led Live SessionsReal-life Case StudiesAssignmentsLifetime AccessExplore Curriculum

Q11. What is ROC curve and what does it represent?

Receiver Operating Characteristic curve (or ROC curve) is a fundamental tool for diagnostic test evaluation and is a plot of the true positive rate (Sensitivity) against the false positive rate (Specificity) for the different possible cut-off points of a diagnostic test.

ROC - Machine Learning Interview Questions - Edureka

ROC – Machine Learning Interview Questions – Edureka

  • It shows the tradeoff between sensitivity and specificity (any increase in sensitivity will be accompanied by a decrease in specificity).
  • The closer the curve follows the left-hand border and then the top border of the ROC space, the more accurate the test.
  • The closer the curve comes to the 45-degree diagonal of the ROC space, the less accurate the test.
  • The slope of the tangent line at a cutpoint gives the likelihood ratio (LR) for that value of the test.
  • The area under the curve is a measure of test accuracy.

Q12. What’s the difference between Type I and Type II error?

Type 1 vs Type 2 Error - Machine Learning Interview Questions - Edureka

Type 1 vs Type 2 Error – Machine Learning Interview Questions – Edureka

Q13. Is it better to have too many false positives or too many false negatives? Explain.

False Negatives vs False Poistives - Machine Learning Interview Questions - Edureka

False Negatives vs False Positives – Machine Learning Interview Questions – Edureka

It depends on the question as well as on the domain for which we are trying to solve the problem. If you’re using Machine Learning in the domain of medical testing, then a false negative is very risky, since the report will not show any health problem when a person is actually unwell. Similarly, if Machine Learning is used in spam detection, then a false positive is very risky because the algorithm may classify an important email as spam.

Q14. Which is more important to you – model accuracy or model performance?

Model Accuracy vs Performance - Machine Learning Interview Questions - Edureka

Model Accuracy vs Performance – Machine Learning Interview Questions – Edureka

Well, you must know that model accuracy is only a subset of model performance. The accuracy of the model and performance of the model are directly proportional and hence better the performance of the model, more accurate are the predictions.

Q15. What is the difference between Gini Impurity and Entropy in a Decision Tree?

  • Gini Impurity and Entropy are the metrics used for deciding how to split a Decision Tree.
  • Gini measurement is the probability of a random sample being classified correctly if you randomly pick a label according to the distribution in the branch.
  • Entropy is a measurement to calculate the lack of information. You calculate the Information Gain (difference in entropies) by making a split. This measure helps to reduce the uncertainty about the output label.

Q16. What is the difference between Entropy and Information Gain?

  • Entropy is an indicator of how messy your data is. It decreases as you reach closer to the leaf node.
  • The Information Gain is based on the decrease in entropy after a dataset is split on an attribute. It keeps on increasing as you reach closer to the leaf node.

Q17. What is Overfitting? And how do you ensure you’re not overfitting with a model?

Over-fitting occurs when a model studies the training data to such an extent that it negatively influences the performance of the model on new data.

This means that the disturbance in the training data is recorded and learned as concepts by the model. But the problem here is that these concepts do not apply to the testing data and negatively impact the model’s ability to classify the new data, hence reducing the accuracy on the testing data.

Three main methods to avoid overfitting:

  • Collect more data so that the model can be trained with varied samples.
  • Use ensembling methods, such as Random Forest. It is based on the idea of bagging, which is used to reduce the variation in the predictions by combining the result of multiple Decision trees on different samples of the data set.
  • Choose the right algorithm.

Q18.Explain Ensemble learning technique in Machine Learning.

 

Ensemble Learning – Machine Learning Interview Questions – Edureka

Ensemble learning is a technique that is used to create multiple Machine Learning models, which are then combined to produce more accurate results. A general Machine Learning model is built by using the entire training data set. However, in Ensemble Learning the training data set is split into multiple subsets, wherein each subset is used to build a separate model. After the models are trained, they are then combined to predict an outcome in such a way that the variance in the output is reduced.

Q19. What is bagging and boosting in Machine Learning?

Bagging & Boosting - Machine Learning Interview Questions - Edureka

Bagging & Boosting – Machine Learning Interview Questions – Edureka

Q20. How would you screen for outliers and what should you do if you find one?

The following methods can be used to screen outliers:

  1. Boxplot: A box plot represents the distribution of the data and its variability. The box plot contains the upper and lower quartiles, so the box basically spans the Inter-Quartile Range (IQR). One of the main reasons why box plots are used is to detect outliers in the data. Since the box plot spans the IQR, it detects the data points that lie outside this range. These data points are nothing but outliers.
  2. Probabilistic and statistical models: Statistical models such as normal distribution and exponential distribution can be used to detect any variations in the distribution of data points. If any data point is found outside the distribution range, it is rendered as an outlier.
  3. Linear models: Linear models such as logistic regression can be trained to flag outliers. In this manner, the model picks up the next outlier it sees.
  4. Proximity-based models: An example of this kind of model is the K-means clustering model wherein, data points form multiple or ‘k’ number of clusters based on features such as similarity or distance. Since similar data points form clusters, the outliers also form their own cluster. In this way, proximity-based models can easily help detect outliers.

How do you handle these outliers?

  • If your data set is huge and rich then you can risk dropping the outliers.
  • However, if your data set is small then you can cap the outliers, by setting a threshold percentile. For example, the data points that are above the 95th percentile can be used to cap the outliers.
  • Lastly, based on the data exploration stage, you can narrow down some rules and impute the outliers based on those business rules.

Q21. What are collinearity and multicollinearity?

  • Collinearity occurs when two predictor variables (e.g., x1 and x2) in a multiple regression have some correlation.
  • Multicollinearity occurs when more than two predictor variables (e.g., x1, x2, and x3) are inter-correlated.

Q22. What do you understand by Eigenvectors and Eigenvalues?

  • Eigenvectors: Eigenvectors are those vectors whose direction remains unchanged even when a linear transformation is performed on them.
  • Eigenvalues: Eigenvalue is the scalar that is used for the transformation of an Eigenvector.
Eigenvalue & Eigenvectors - Machine Learning Interview Questions - Edureka

Eigenvalue & Eigenvectors – Machine Learning Interview Questions – Edureka

In the above example, 3 is an Eigenvalue, with the original vector in the multiplication problem being an eigenvector.

Data Science Training

PYTHON CERTIFICATION TRAINING FOR DATA SCIENCEPython Certification Training for Data ScienceReviews 5(50328)PYTHON PROGRAMMING CERTIFICATION COURSEPython Programming Certification CourseReviews 5(9029)MACHINE LEARNING CERTIFICATION TRAINING USING PYTHONMachine Learning Certification Training using PythonReviews 5(5541)DATA SCIENCE CERTIFICATION COURSE USING RData Science Certification Course using RReviews 5(34627)DATA ANALYTICS WITH R CERTIFICATION TRAININGData Analytics with R Certification TrainingReviews 5(21738)STATISTICS ESSENTIALS FOR ANALYTICSStatistics Essentials for AnalyticsReviews 5(4687)SAS TRAINING AND CERTIFICATIONSAS Training and CertificationReviews 5(3900)ANALYTICS FOR RETAIL BANKSAnalytics for Retail BanksReviews 5(750)DECISION TREE MODELING USING R CERTIFICATION TRAININGDecision Tree Modeling Using R Certification TrainingReviews 5(1369)Next

The Eigenvector of a square matrix A is a nonzero vector x such that for some number λ, we have the following:

Ax = λx, 

where λ is an Eigenvalue
So, in our example, λ = 3 and X = [1 1 2]

Q23. What is A/B Testing?

  • A/B is Statistical hypothesis testing for randomized experiment with two variables A and B. It is used to compare two models that use different predictor variables in order to check which variable fits best for a given sample of data.
  • Consider a scenario where you’ve created two models (using different predictor variables) that can be used to recommend products for an e-commerce platform.
  • A/B Testing can be used to compare these two models to check which one best recommends products to a customer.
A/B Testing - Machine Learning Interview Questions - Edureka

A/B Testing – Machine Learning Interview Questions – Edureka

Q24. What is Cluster Sampling?

  • It is a process of randomly selecting intact groups within a defined population, sharing similar characteristics.
  • Cluster Sample is a probability sample where each sampling unit is a collection or cluster of elements.
  • For example, if you’re clustering the total number of managers in a set of companies, in that case, managers (samples) will represent elements and companies will represent clusters.

Q25. Running a binary classification tree algorithm is quite easy. But do you know how the tree decides on which variable to split at the root node and its succeeding child nodes?

  • Measures such as, Gini Index and Entropy can be used to decide which variable is best fitted for splitting the Decision Tree at the root node.
  • We can calculate Gini as following:
    Calculate Gini for sub-nodes, using the formula – sum of square of probability for success and failure (p^2+q^2).
  • Calculate Gini for split using weighted Gini score of each node of that split
  • Entropy is the measure of impurity or randomness in the data, (for binary class):
Entropy - Machine Learning Interview Questions - Edureka

Here p and q is the probability of success and failure respectively in that node.

  • Entropy is zero when a node is homogeneous and is maximum when both the classes are present in a node at 50% – 50%. To sum it up, the entropy must be as low as possible in order to decide whether or not a variable is suitable as the root node.

Machine Learning With Python Questions

This set of Machine Learning interview questions deal with Python related Machine Learning questions.

Q1. Name a few libraries in Python used for Data Analysis and Scientific Computations.

Here is a list of Python libraries mainly used for Data Analysis:

  • NumPy
  • SciPy
  • Pandas
  • SciKit
  • Matplotlib
  • Seaborn
  • Bokeh

Q2. Which library would you prefer for plotting in Python language: Seaborn or Matplotlib or Bokeh?

Python Libraries - Machine Learning Interview Questions - Edureka

Python Libraries – Machine Learning Interview Questions – Edureka

It depends on the visualization you’re trying to achieve. Each of these libraries is used for a specific purpose:

  • Matplotlib: Used for basic plotting like bars, pies, lines, scatter plots, etc
  • Seaborn: Is built on top of Matplotlib and Pandas to ease data plotting. It is used for statistical visualizations like creating heatmaps or showing the distribution of your data
  • Bokeh: Used for interactive visualization. In case your data is too complex and you haven’t found any “message” in the data, then use Bokeh to create interactive visualizations that will allow your viewers to explore the data themselves

Q3. How are NumPy and SciPy related?

  • NumPy is part of SciPy.
  • NumPy defines arrays along with some basic numerical functions like indexing, sorting, reshaping, etc.
  • SciPy implements computations such as numerical integration, optimization and machine learning using NumPy’s functionality.

Q4. What is the main difference between a Pandas series and a single-column DataFrame in Python?

Pandas Series vs DataFrame - Machine Learning Interview Questions - Edureka

Pandas Series vs DataFrame – Machine Learning Interview Questions – Edureka

Q5. How can you handle duplicate values in a dataset for a variable in Python?

Consider the following Python code:

1234567bill_data=pd.read_csv("datasetsTelecom Data AnalysisBill.csv")bill_data.shape#Identify duplicates records in the dataDupes = bill_data.duplicated()sum(dupes)#Removing Duplicatesbill_data_uniq = bill_data.drop_duplicates()

Q6. Write a basic Machine Learning program to check the accuracy of a model,  by importing any dataset using any classifier?

123456789101112131415161718#importing datasetimport sklearnfrom sklearn import datasetsiris = datasets.load_iris()X = iris.dataY = iris.target #splitting the datasetfrom sklearn.cross_validation import train_test_splitX_train, Y_train, X_test, Y_test = train_test_split(X,Y, test_size = 0.5) #Selecting Classifiermy_classifier = tree.DecisionTreeClassifier()My_classifier.fit(X_train, Y_train)predictions = my_classifier(X_test)#check accuracyFrom sklear.metrics import accuracy_scoreprint accuracy_score(y_test, predictions)

Machine Learning Scenario Based Questions

This set of Machine Learning interview questions deal with scenario-based Machine Learning questions.

Q1. You are given a data set consisting of variables having more than 30% missing values? Let’s say, out of 50 variables, 8 variables have missing values higher than 30%. How will you deal with them?

  • Assign a unique category to the missing values, who knows the missing values might uncover some trend.
  • We can remove them blatantly.
  • Or, we can sensibly check their distribution with the target variable, and if found any pattern we’ll keep those missing values and assign them a new category while removing others.

Q2. Write an SQL query that makes recommendations using the pages that your friends liked. Assume you have two tables: a two-column table of users and their friends, and a two-column table of users and the pages they liked. It should not recommend pages you already like.

12345SELECT f.user_id, l.page_idFROM friend f JOIN like lON f.friend_id = l.user_idWHERE l.page_id NOT IN (SELECT page_id FROM likeWHERE user_id = f.user_id)

Q3. There’s a game where you are asked to roll two fair six-sided dice. If the sum of the values on the dice equals seven, then you win $21. However, you must pay $5 to play each time you roll both dice. Do you play this game? And in the follow-up: If he plays 6 times what is the probability of making money from this game?

  • The first condition states that if the sum of the values on the 2 dices is equal to 7, then you win $21. But for all the other cases you must pay $5.
  • First, let’s calculate the number of possible cases. Since we have two 6-sided dices, the total number of cases => 6*6 = 36.
  • Out of 36 cases, we must calculate the number of cases that produces a sum of 7 (in such a way that the sum of the values on the 2 dices is equal to 7)
  • Possible combinations that produce a sum of 7 is, (1,6), (2,5), (3,4), (4,3), (5,2) and (6,1). All these  6 combinations generate a sum of 7.
  • This means that out of 36 chances, only 6 will produce a sum of 7. On taking the ratio, we get: 6/36 = 1/6
  • So this suggests that we have a chance of winning $21, once in 6 games.
  • So to answer the question if a person plays 6 times, he will win one game of $21, whereas for the other 5 games he will have to pay $5 each, which is $25 for all five games. Therefore, he will face a loss because he wins $21 but ends up paying $25.

Q4. We have two options for serving ads within Newsfeed:
1 – out of every 25 stories, one will be an ad
2 – every story has a 4% chance of being an ad

For each option, what is the expected number of ads shown in 100 news stories?
If we go with option 2, what is the chance a user will be shown only a single ad in 100 stories? What about no ads at all?

  • The expected number of ads shown in 100 new stories for option 1 is equal to 4 (100/25 = 4).
  • Similarly, for option 2, the expected number of ads shown in 100 new stories is also equal to 4 (4/100 = 1/25 which suggests that one out of every 25 stories will be an ad, therefore in 100 new stories there will be 4 ads)
  • Therefore for each option, the total number of ads shown in 100 new stories is 4.
  • The second part of the question can be solved by using Binomial distribution. Binomial distribution takes three parameters:
    • The probability of success and failure, which in our case is 4%.
    • The total number of cases, which is 100 in our case.
    • The probability of the outcome, which is a chance that a user will be shown only a single ad in 100 stories
  • p(single ad) = (0.96)^99*(0.04)^1

(note: here 0.96 denotes the chance of not seeing an ad in 100 stories, 99 denotes the possibility of seeing only 1 ad, 0.04 is the probability of seeing an ad once in 100 stories )

  • In total, there are 100 positions for the ad. Therefore, 100 * p(single ad) = 7.03%

Q5. How would you predict who will renew their subscription next month? What data would you need to solve this? What analysis would you do? Would you build predictive models? If so, which algorithms?

  • Let’s assume that we’re trying to predict renewal rate for Netflix subscription. So our problem statement is to predict which users will renew their subscription plan for the next month.
  • Next, we must understand the data that is needed to solve this problem. In this case, we need to check the number of hours the channel is active for each household, the number of adults in the household, number of kids, which channels are streamed the most, how much time is spent on each channel, how much has the watch rate varied from last month, etc. Such data is needed to predict whether or not a person will continue the subscription for the upcoming month.
  • After collecting this data, it is important that you find patterns and correlations. For example, we know that if a household has kids, then they are more likely to subscribe. Similarly, by studying the watch rate of the previous month, you can predict whether a person is still interested in a subscription. Such trends must be studied.
  • The next step is analysis. For this kind of problem statement, you must use a classification algorithm that classifies customers into 2 groups:
    • Customers who are likely to subscribe next month
    • Customers who are not likely to subscribe next month
  • Would you build predictive models? Yes, in order to achieve this you must build a predictive model that classifies the customers into 2 classes like mentioned above.
  • Which algorithms to choose? You can choose classification algorithms such as Logistic Regression, Random Forest, Support Vector Machine, etc.
  • Once you’ve opted the right algorithm, you must perform model evaluation to calculate the efficiency of the algorithm. This is followed by deployment.

Q6. How do you map nicknames (Pete, Andy, Nick, Rob, etc) to real names?

  • This problem can be solved in n number of ways. Let’s assume that you’re given a data set containing 1000s of twitter interactions. You will begin by studying the relationship between two people by carefully analyzing the words used in the tweets.
  • This kind of problem statement can be solved by implementing Text Mining using Natural Language Processing techniques, wherein each word in a sentence is broken down and co-relations between various words are found.
  • NLP is actively used in understanding customer feedback, performing sentimental analysis on Twitter and Facebook. Thus, one of the ways to solve this problem is through Text Mining and Natural Language Processing techniques.

Q7. A jar has 1000 coins, of which 999 are fair and 1 is double headed. Pick a coin at random, and toss it 10 times. Given that you see 10 heads, what is the probability that the next toss of that coin is also a head?

  • There are two ways of choosing a coin. One is to pick a fair coin and the other is to pick the one with two heads.
  • Probability of selecting fair coin = 999/1000 = 0.999
    Probability of selecting unfair coin = 1/1000 = 0.001
  • Selecting 10 heads in a row = Selecting fair coin * Getting 10 heads + Selecting an unfair coin
  • P (A) = 0.999 * (1/2)^10 = 0.999 * (1/1024) = 0.000976
    P (B) = 0.001 * 1 = 0.001
    P( A / A + B ) = 0.000976 / (0.000976 + 0.001) = 0.4939
    P( B / A + B ) = 0.001 / 0.001976 = 0.5061
  • Probability of selecting another head = P(A/A+B) * 0.5 + P(B/A+B) * 1 = 0.4939 * 0.5 + 0.5061 = 0.7531

Q8. Suppose you are given a data set which has missing values spread along 1 standard deviation from the median. What percentage of data would remain unaffected and Why?

Since the data is spread across the median, let’s assume it’s a normal distribution.
As you know, in a normal distribution, ~68% of the data lies in 1 standard deviation from mean (or mode, median), which leaves ~32% of the data unaffected. Therefore, ~32% of the data would remain unaffected by missing values.

Q9. You are given a cancer detection data set. Let’s suppose when you build a classification model you achieved an accuracy of 96%. Why shouldn’t you be happy with your model performance? What can you do about it?

You can do the following:

  • Add more data
  • Treat missing outlier values
  • Feature Engineering
  • Feature Selection
  • Multiple Algorithms
  • Algorithm Tuning
  • Ensemble Method
  • Cross-Validation

Q10. You are working on a time series data set. Your manager has asked you to build a high accuracy model. You start with the decision tree algorithm since you know it works fairly well on all kinds of data. Later, you tried a time series regression model and got higher accuracy than the decision tree model. Can this happen? Why?

  • Time series data is based on linearity while a decision tree algorithm is known to work best to detect non-linear interactions
  • Decision tree fails to provide robust predictions. Why?
    • The reason is that it couldn’t map the linear relationship as good as a regression model did.
    • We also know that a linear regression model can provide a robust prediction only if the data set satisfies its linearity assumptions.

Q11. Suppose you found that your model is suffering from low bias and high variance. Which algorithm you think could tackle this situation and Why?

Type 1: How to tackle high variance?

  • Low bias occurs when the model’s predicted values are near to actual values.
  • In this case, we can use the bagging algorithm (eg: Random Forest) to tackle high variance problem.
  • Bagging algorithm will divide the data set into its subsets with repeated randomized sampling.
  • Once divided, these samples can be used to generate a set of models using a single learning algorithm. Later, the model predictions are combined using voting (classification) or averaging (regression).

Type 2: How to tackle high variance?

  • Lower the model complexity by using regularization technique, where higher model coefficients get penalized.
  • You can also use top n features from variable importance chart. It might be possible that with all the variable in the data set, the algorithm is facing difficulty in finding the meaningful signal.

Q12. You are given a data set. The data set contains many variables, some of which are highly correlated and you know about it. Your manager has asked you to run PCA. Would you remove correlated variables first? Why?

Possibly, you might get tempted to say no, but that would be incorrect.
Discarding correlated variables will have a substantial effect on PCA because, in the presence of correlated variables, the variance explained by a particular component gets inflated.

Q13. You are asked to build a multiple regression model but your model R² isn’t as good as you wanted. For improvement, you remove the intercept term now your model R² becomes 0.8 from 0.3. Is it possible? How?

Yes, it is possible.

  • The intercept term refers to model prediction without any independent variable or in other words, mean prediction
    R² = 1 – ∑(Y – Y´)²/∑(Y – Ymean)² where Y´ is the predicted value.
  • In the presence of the intercept term, R² value will evaluate your model with respect to the mean model.
  • In the absence of the intercept term (Ymean), the model can make no such evaluation,
  • With large denominator,
    Value of ∑(Y – Y´)²/∑(Y)² equation becomes smaller than actual, thereby resulting in a higher value of R².

Q14. You’re asked to build a random forest model with 10000 trees. During its training, you got training error as 0.00. But, on testing the validation error was 34.23. What is going on? Haven’t you trained your model perfectly?

  • The model is overfitting the data.
  • Training error of 0.00 means that the classifier has mimicked the training data patterns to an extent.
  • But when this classifier runs on the unseen sample, it was not able to find those patterns and returned the predictions with more number of errors.
  • In Random Forest, it usually happens when we use a larger number of trees than necessary. Hence, to avoid such situations, we should tune the number of trees using cross-validation.

Q15. ‘People who bought this also bought…’ recommendations seen on Amazon is based on which algorithm?

E-commerce websites like Amazon make use of Machine Learning to recommend products to their customers. The basic idea of this kind of recommendation comes from collaborative filtering. Collaborative filtering is the process of comparing users with similar shopping behaviors in order to recommend products to a new user with similar shopping behavior.Machine Learning Certification Training using PythonWeekday / Weekend BatchesSee Batch Details

Collaborative Filtering - Machine Learning Interview Questions - Edureka

Collaborative Filtering – Machine Learning Interview Questions – Edureka

To better understand this, let’s look at an example. Let’s say a user A who is a sports enthusiast bought, pizza, pasta, and a coke.  Now a couple of weeks later, another user B who rides a bicycle buys pizza and pasta. He does not buy the coke, but Amazon recommends a bottle of coke to user B since his shopping behaviors and his lifestyle is quite similar to user A. This is how collaborative filtering works.

So these are the most frequently asked questions in a Machine Learning Interview. However, if you wish to brush up more on your knowledge, you can go through these blogs:

  1. Machine Learning Tutorial for Beginners
  2. Top 10 Applications of Machine Learning: Machine Learning Applications in Daily Life
  3. Machine Learning Algorithms

With this, we come to an end of this blog. I hope these Machine Learning Interview Questions will help you ace your Machine Learning Interview.

If you want to become a successful Machine Learning Engineer, you can take up the Machine Learning Certification Training using Python from Edureka. This program exposes you to concepts of Statistics, Time Series and different classes of machine learning algorithms like supervised, unsupervised and reinforcement algorithms. It will make you proficient in various Machine Learning algorithms such as Regression, Clustering, Decision Trees, Random Forest, Naïve Baye, and Q-Learning.

Recommended videos for you

 

Data Science : Make Smarter Business Decisions

Watch Now

The Whys and Hows of Predictive Modelling-I

Watch Now

Sentiment Analysis In Retail Domain

Watch Now

The Whys and Hows of Predictive Modeling-II

Watch Now

Python Programming – Learn Python Programming From Scratch

Watch Now

Python Classes – Python Programming Tutorial

Watch Now

Python List, Tuple, String, Set And Dictonary – Python Sequences

Watch Now

Introduction to Business Analytics with R

Watch Now

Web Scraping And Analytics With Python

Watch Now

Application of Clustering in Data Science Using Real-Time Examples

Watch Now

Python for Big Data Analytics

Watch Now

Business Analytics with R

Watch Now

Know The Science Behind Product Recommendation With R Programming

Watch Now

Python Numpy Tutorial – Arrays In Python

Watch Now

Mastering Python : An Excellent tool for Web Scraping and Data Analysis

Watch Now

Python Tutorial – All You Need To Know In Python Programming

Watch Now

Python Loops – While, For and Nested Loops in Python Programming

Watch Now

Business Analytics Decision Tree in R

Watch Now

Android Development : Using Android 5.0 Lollipop

Watch Now

Machine Learning With Python – Python Machine Learning Tutorial

Watch Now

Recommended blogs for you

 

Everything You Need To Know About Hash In Python

Read Article

Python vs C: Know what are the differences

Read Article

Python Decorator Tutorial : How To Use Decorators In Python

Read Article

Python vs C++: Know what are the differences

Read Article

A Comprehensive Guide To R For Data Science

Read Article

What is Business Analytics? All you Need to Know

Read Article

Python time sleep() – One Stop Solution for time.sleep() Method

Read Article

If Else In Python With Examples : Everything You Need To Know

Read Article

Exceptions in Python

Read Article

What is Queue Data Structure In Python?

Read Article

Top 10 Machine Learning Frameworks You Need to Know

Read Article

Types of Sentiment Analysis

Read Article

Install Python On Windows – Python 3.X Installation Guide

Read Article

How to Implement Power Function in Python

Read Article

Python Modules- All You Need To know

Read Article

Learn How To Make Simple Mobile Applications Using This Kivy Tutorial In Python

Read Article

Introduction To File Handling In Python

Read Article

Different Job Titles for Data Scientists

Read Article

String Trimming in Python: All you Need to Know

Read Article

Modeling Techniques in Business Analytics with R

Read ArticleComments0 Comments

Trending Courses in Data Science

Python Certification Training for Data Scienc …51k Enrolled LearnersWeekend/WeekdayLive ClassReviews 5 (20150)Python Programming Certification Course10k Enrolled LearnersWeekendLive ClassReviews 5 (3650)Machine Learning Certification Training using …6k Enrolled LearnersWeekendLive ClassReviews 5 (2250)Data Science Certification Course using R35k Enrolled LearnersWeekendLive ClassReviews 5 (13900)Data Analytics with R Certification Training22k Enrolled LearnersWeekendLive ClassReviews 5 (8700)Statistics Essentials for Analytics5k Enrolled LearnersWeekend/WeekdaySelf PacedReviews 5 (1900)SAS Training and Certification4k Enrolled LearnersWeekendLive ClassReviews 5 (1600)Analytics for Retail Banks1k Enrolled LearnersWeekend/WeekdaySelf PacedReviews 5 (300)Decision Tree Modeling Using R Certification …2k Enrolled LearnersWeekend/WeekdaySelf PacedReviews 5 (550)Advanced Predictive Modelling in R Certificat …4k Enrolled LearnersWeekendSelf PacedReviews 4 (1400)

Browse Categories

Artificial IntelligenceBI and VisualizationBig DataBlockchainCloud ComputingCyber SecurityData Warehousing and ETLDatabasesDevOpsDigital MarketingFront End Web DevelopmentMobile DevelopmentOperating SystemsProgramming & FrameworksProject Management and MethodologiesRobotic Process AutomationSoftware TestingSystems & Architecture

Subscribe to our Newsletter, and get personalized recommendations.

Google

 Sign up with Google 

facebook

 Signup with Facebook

Already have an account? Sign in

TRENDING CERTIFICATION COURSES

TRENDING MASTERS COURSES

COMPANY

WORK WITH US

DOWNLOAD APP

 

google_playstore

CATEGORIES

TRENDING BLOG ARTICLES

© 2019 Brain4ce Education Solutions Pvt. Ltd. All rights Reserved. Terms & ConditionsLegal & Privacy   “PMP®”,”PMI®”, “PMI-ACP®” and “PMBOK®” are registered marks of the Project Management Institute, Inc. MongoDB®, Mongo and the leaf logo are the registered trademarks of MongoDB, Inc.

Scientific citation

From Wikipedia, the free encyclopediaJump to navigationJump to searchFor Wikipedia’s guide to referencing scientific and mathematical articles, see Wikipedia:Scientific citation guidelines

 This article needs additional citations for verification. Please help improve this article by adding citations to reliable sources. Unsourced material may be challenged and removed.
Find sources: “Scientific citation” – news · newspapers · books · scholar · JSTOR (August 2007) (Learn how and when to remove this template message)

Reference section in scientific paper.

Scientific citation is providing detailed reference in a scientific publication, typically a paper or book, to previous published (or occasionally private) communications which have a bearing on the subject of the new publication. The purpose of citations in original work is to allow readers of the paper to refer to cited work to assist them in judging the new work, source background information vital for future development, and acknowledge the contributions of earlier workers. Citations in, say, a review paper bring together many sources, often recent, in one place.

To a considerable extent the quality of work, in the absence of other criteria, is judged on the number of citations received, adjusting for the volume of work in the relevant topic.[citation needed] While this is not necessarily a reliable measure, counting citations is trivially easy; judging the merit of complex work can be very difficult.[citation needed]

Previous work may be cited regarding experimental procedures, apparatus, goals, previous theoretical results upon which the new work builds, theses, and so on. Typically such citations establish the general framework of influences and the mindset of research, and especially as “part of what science” it is, and to help determine who conducts the peer review.[citation needed]

Disciplined citation of prior works in mathematics and science is known at least as far back as Euclid.[citation needed] Late in the first millennium, Islamic scholars developed their practice of isnad, or “backing”, which established the validity of sayings of Muhammad in the hadith.[citation needed] The Asharite school of early Muslim philosophy extended this into fiqh or jurisprudence, while the Mutazilite school used the traditional methods and applied them to science.[citation needed]

In some form, then, achieving authority for new work by citing accepted authorities is a near-universal idea among the peoples of the Mediterranean, whose educated people were exposed to one or other of these practices well before the European Renaissance and the emergence of the formal scientific method.[citation needed]

Contents

Patent references[edit]

In patent law the citation of previous works, or prior art, helps establish the uniqueness of the invention being described. The focus in this practice is to claim originality for commercial purposes, and so the author is motivated to avoid citing works that cast doubt on its originality. Thus this does not appear to be “scientific” citation. Inventors and lawyers have a legal obligation to cite all relevant art; not to do so risks invalidating the patent.[citation needed] The patent examiner is obliged to list all further prior art found in searches.[citation needed]

Citation frequency[edit]

Modern scientists are sometimes judged by the number of times their work is cited by others—this is actually a key indicator of the relative importance of a work in science. Accordingly, individual scientists are motivated to have their own work cited early and often and as widely as possible, but all other scientists are motivated to eliminate unnecessary citations so as not to devalue this means of judgment.[citation needed] A formal citation index tracks which referred and reviewed papers have referred which other such papers. Baruch Lev and other advocates of accounting reform consider the number of times a patent is cited to be a significant metric of its quality, and thus of innovation.[citation needed]

See also[edit]

References[edit]

Further reading[edit]

External links[edit]

Categories

Navigation menu

Search

Interaction

Tools

Print/export

Languages

Edit links

  •  
  •  

 

44 Comments

  1. I savor, lead to I found just what I used to be taking a look for. You’ve ended my 4 day lengthy hunt! God Bless you man. Have a great day. Bye

    Like

  2. Hello, I believe your website may be having internet browser compatibility issues.

    When I look at your website in Safari, it looks fine but when opening in I.E., it’s
    got some overlapping issues. I merely wanted to provide you
    with a quick heads up! Other than that, excellent website!

    Like

  3. Wonderful items from you, mɑn. I have take into accout your stᥙff previous
    to and you’re simply extremely great. I actually like wһat
    you have got riցht here, certainly like whɑt you are stating and the way wherein you say it.

    You make it еnjoyable and you ѕtill care f᧐r to kеep it smart.
    I ϲan’t wait to read far more from you. That is really a tremendous wеbsite.

    Like

  4. I’m truly enjoying the design and layout of your website. It’s a very easy on the eyes which makes it much more pleasant for me to come here and visit more often. Did you hire out a developer to create your theme? Superb work!

    Like

  5. I would like to thank you for the efforts you have put in writing this website. I’m hoping to see the same high-grade content by you in the future as well. In truth, your creative writing abilities has inspired me to get my own site now 😉

    Like

  6. Does your site have a contact page? I’m having problems locating it but, I’d like to shoot you an email. I’ve got some suggestions for your blog you might be interested in hearing. Either way, great blog and I look forward to seeing it grow over time.

    Like

  7. Soy de búsqueda inesperada atló la oportunidad de conocer gente interesante para hacer un flujo, una conversación con otras personas que llenan los mismos
    gustos. Debes estudiar siempre la celebración y dedicación a los hombres,
    algo que se vaya más facilidad para ayudarte a mantenerte gente con la información y hacia tu pareja.
    Por ejemplo, cuando los usuarios contriban a nuestros propios sitios de
    citas facilita las cosas por naturaleza, o cuando no contaba el lenguaje corporal de hombres, sobre
    todo, si contraitos. Me ose luchida por mi persona. Pero aceptérraste…
    No hay ninguna estabilidad para saber a quién detallada de nuestros datos..

    Like

  8. Thanks for a marvelous posting! I definitely enjoyed reading it, you will be a great author.I will remember to bookmark your blog and definitely will come back later on. I want to encourage continue your great posts, have a nice weekend!

    Like

  9. Excellent website you have here but I was wanting to know if you knew of any message boards that cover the same topics discussed in this article? I’d really love to be a part of community where I can get advice from other experienced individuals that share the same interest. If you have any suggestions, please let me know. Thanks!

    Like

  10. Thanks for your marvelous posting! I quite enjoyed reading it, you can be a great author.I will remember to bookmark your blog and may come back down the road. I want to encourage one to continue your great posts, have a nice morning!

    Like

  11. It’s really a nice and helpful piece of information. I am glad that you shared this useful info with us. Please keep us informed like this. Thanks for sharing.

    Like

  12. Youre so cool! I dont suppose Ive read anything like this before. So nice to find somebody with some original thoughts on this subject. realy thank you for starting this up. this website is something that is needed on the web, someone with a little originality. useful job for bringing something new to the internet!

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s