10 Machine Learning Interview Questions - ANSWERED

We cover 10 machine learning interview questions. Have you had interesting interview experiences you'd like to share? Leave them in the comments!
My channel: / codeemporium
To submit your video to CS Dojo Community, please use this link: csdojo.io/enter
REFERENCES:
[1] Interview Questions: www.springboard.com/blog/mach...
[2] Generative Vs Discriminative: stats.stackexchange.com/quest...
[3]: What is Bayes Rule: www.askamathematician.com/2011...
[4] The likelihood function is not PDF: stats.stackexchange.com/quest...
[5] Cross Validation for time series: robjhyndman.com/hyndsight/tscv/
[6] SMOTE: jair.org/index.php/jair/artic...
[7] Combating Imbalanced Data: www.analyticsvidhya.com/blog/...
[8] Understanding ROC: towardsdatascience.com/unders...
[9] Publicly available datasets: www.springboard.com/blog/free...
[10] Probability Vs Likelihood: stats.stackexchange.com/quest...

[@CSDojoCommunity]

Happy new year everyone! This is the first video on this channel :) And if you’re interested in getting your video or article featured, please take a look at this article: medium.com/p/785731d18a66/

[@emilfilipov169]

I've never done winsorizing. I'll definitely try it for my next experiment!

[@shobhamourya8396]

#1 Probability is given the distribution what is the chance of occurrence of an event whereas Likelihood is given the event what is it chance of occurrence for any distribution, thus finding the maximum likelihood by comparing different distributions.

[@MrLlangz]

I like that you've put references for further reading,good job

[@CodeEmporium]

Happy 2019 everyone! Honored to have the first video on this channel! And hoping to make more here in the future. If you are interested in learning more about machine learning, deep learning, & data sciences, stop by channel to say hi ;)

[@sairamreddypulikanti1153]

CodeEmporium Could you create a playlist on your channel and provide your intuition to interview questions every week(at least 5 questions a week). No one is posting such clear explanations. Love from Germany!

[@CodeEmporium]

@@sairamreddypulikanti1153 I was thinking of something similar! Imma do that soon

[@pixel7038]

@@CodeEmporium What's the difference between both of your channels? It feels redundant to subscribe to both

[@shaunpx1]

Awesome video! For questions #1 an easier way to explain 🤔probability vs likelihood is. Probability is based on the outcome of some sample being applied to a distribution. Whereas, Likleyhood is finding the best distribution of data given some value.

[@AllMightGaming-AMG]

Agreed, this is the more general answer. In the context of machine learning, instead of the distribution, we are interested in the model parameter theta.

[@AhmetBurakAkay]

Best video I have seen so far. Thanks.

[@deeganishot]

Just wanted to let you know I really enjoyed your video. Paused it right after you asked the question to give it a go. Didn't get that many right myself, but am taking it as a learning experience. 10 questions closer in preparation :-)

[@CodeEmporium]

That's great Rob! Glad to know you interacted with the video, and learned some thing new :)

[@MultiAryanraj]

Thanks alot. This video really helped to get the idea of the type of questions and topics in an interview. Can you post a video mentioning how to prepare for an ML/DS interview(because there are so many things to study)? How to get noticed by recruiter and getting selected for an interview? Getting the interview call is very hard nowadays even after applying in so many organizations.

[@edheinbockel9653]

Great info, thank you for sharing!

[@WalkAloneLive]

Great video, thanks!!

[@vinayreddy8683]

The way you've changed your accent within short time is amazing.

[@neilpradhan1312]

Very good ! Thanks for sharing

[@eh0011he]

awesome video!! thx

[@muskduh]

Thanks for the video

[@butoyighyslain171]

yes please create a data science/ML interviews questions!

[@sridharar4455]

Great Content

[@neilrodrigues2710]

equation at 3:53 is wrong

[@2odhwka]

Thanks alot.

[@msnhao]

4:00 Your baye's theorem should be divided by P(E).

[@torch_9001]

Subbed ✔

[@CodeEmporium]

Welcome homie! And Thanks!

[@drewgrant1605]

@8:45 *SMOTE=Synthetic Minority Oversampling Technique

[@botfactory1510]

nice video :)

[@dipenduchoudhury9984]

Simplest meaning of likelihood and probability are below... Likelihood where we get the probability when label is occurred and this probability score store into memory. .................... And probability is used in testing phase. We are trying to predict the probability when features are true

[@matthewyeseta4704]

Working on my Master of Data Science and this video is very nice for review

[@juanete69]

Your first formula in the Bayes section is not right, the denominator should be P(E).

[@Reflectives]

Why us? You pay enough, I can do the job

[@raffaelepiccini3405]

I think that your answers are a bit overcomplicated for an interview, they are well done but not very good in an interview setting

[@haSHAH1]

Why is it so "Math-y and Stats-y"?

[@VikasKcgeek]

What kya hai ye. Copy Paste

[@tushar12345]

its like freeCodeCamp, where you can submit your own video to be featured in this channel

[@pratikdeshmukh4268]

unnecessarily complicated

[@haSHAH1]

It seems that you could be a little more polite!