Machine Learning Tutorial Python - 20: Bias vs Variance In Machine Learning

[@codebasics]

Check out our premium machine learning course with 2 Industry projects: codebasics.io/courses/machine-learning-for-data-science-beginners-to-advanced

[@kent2super9]

So thankful for the efforts. I am taking a AIML certification and key concepts seem to be missed. I am literally using your videos in parallel with the class to close gaps and improve my understanding. I teach SAP courses and Power BI, so I understand the time it takes to create quality training videos. The ability to take complex subjects and explain them in such a way my grandpa could understand it, is a skill. Hats off to you sir.

[@codebasics]

Victor thank you for your kind words of appreciation 🙏

[@tesfayesusyimenu3292]

You deserve an award for making concepts clearer!

[@daniellemccorkle6917]

This entire series is fabulous and super relevant!! Thank you for these videos, greatly appreciated!!

[@Marshall_Mohammed]

Hats Off to you and your efforts, you are simplifying ML for this generation. Your way of teaching is irreplaceable!❤

[@techsavy5669]

Bullseye for bulls eye diagram explanation. Just awesome.

[@codebasics]

Glad it was helpful!

[@prathameshmore5262]

great presentation i ever saw . I can clearly see that how test error depends on selection of train datapoints . Thankyou sir

[@ontreprenor]

Wow. Such an amazing explanation. I watched 3 videos before yours and none were as explanatory as yours!!!

[@zehaia]

That is the best explanation of bias/variance tradeoff on KZbin. I wish you will make a series on advanced level machine and deep learning. Especially about the underlying math.

[@cityrunner-x3x]

This is very genius example of underfitting and overfitting. Love it and thanks, haha.

[@dataguy7013]

Best , detailed and intuitive example that is TRULY understandable. Never seen something some like this before. Thank you!!!

[@artiverma1402]

Hats off to you to explain in such a simple way

[@bakerkar]

superb video. Far superior to the Lectures by IIT profs on this subject. Great work and wishing yoy great success in the future

[@redzone8115]

Thanks for your Ameging Video, this video clear my concept about Bias and Variance...

[@jeekakrishna]

Sir you continue with this please

[@valli9626]

A very clear explanation!!

[@moahaimen]

Greatest Teacher

[@danielasefa8087]

thank you so much for the constructive and clear explanation

[@vinayak254]

Thank you sir for teaching everything simple. It is easy to remember also. Great!!

[@chuckyneoable]

Thank you so much for such a clear illustration and explanation

[@madhupincha7898]

Great explanation in layman words

[@vgreddysaragada]

Super description..Thank you

[@prayagrajchaudhary2270]

This is a very owsome course designed by you sir. Thanks for your efforts.

[@its_kumar]

You are always a savior 🙏

[@codebasics]

Thanks Kumar, hope you are doing well my friend

[@its_kumar]

@@codebasics yes sir, I'm good ☺️

[@abdolrezamohseni9787]

Very great explanation. Thanks so much for that

[@nastaran1010]

Best training. thanks

[@adityaaggarwal424]

Well explained! Thanks for the effort Sir!!

[@MK-yj7pn]

pretty good explanation.

[@sakshigaikwad8711]

Nice explanation

[@aibi1910]

Awesome Explanation :)

[@gujrathisiddhant1112]

Great😇

[@rukhsananazz2747]

clearly explained

[@HT-xt4cn]

Thanks for the video. I have a question: Why should we be concerned if our model produces high bias? Surely it is the test set, i.e. the variance, that should concern us, is it not?

[@Dyslexic_Neuron]

Great explanation....better than statquest

[@codebasics]

I am happy this was helpful to you.

[@jayshreedonga2833]

Thanks sir

[@write2ruby]

Very Nice

[@darvishdavis159]

Salute you sir,

[@dr.sumitdesai8458]

Superb explaination

[@codebasics]

Glad you liked it

[@mohammedalatrash1973]

awesome thanks

[@kmnm9463]

Hi, Great content. The best in YT on bias and variance. I have a doubt - from 07:35 - 07:40 in the video, while we are looking at an ideal model, there are two curves which have been shown - meaning these are two different models. I thought we are looking into finding a single model which has a balanced fit. While we are varying the training dataset, the model also is changed. I feel it should the same curve for different training datasets. Regards, Krish

[@jamalnuman]

great

[@Mary-gl4lz]

Hello Sir If for bias_var_decomp method if we are not mentioning loss, by default what will it take as loss? loss, bias, var =bias_variance_decomp(model,X_train.values, y_trainnp, X_test.values, y_testnp)

[@jcv71]

Are this Machien Learning videos in a playlist, I can't find it on your playlists

[@RajkumarDarbar]

best explanation !!

[@parth_suthar]

Hi , sir i am parth and iam from Dakor and. i am studies in adit anand

[@axolotl7701]

Variance = Model Training Error (Sample Value) Bias = Model Test Error (The True Value)

[@nyozha7154]

He said exactly the contrary, which is weird so I'm confused.

[@SanjanaGupta-jt1so]

sir in second case there is train error is 43 and 2nd time train error is 41 so there is not much difference then how it become high bias?

[@nikssluv]

Hello sir, I did my graduation in mechanical in 2013 . I Have 6 year of career gap. From last 2 year i m working as software engineer. Now i m thinking to PG diploma in Data science from coursera. IN NEXT YEAR After completing the diploma course in data science. I am thinking to apply for master in Germany in Data science. WHAT IS THE CHANCE TO SELECT IN MASTER COURSE. Kindly suggest me some right career path. IS IT POSSIBLE to land in masters courses if i have 6 year of carrers gap along with 2 year of experience

[@nikssluv]

Sir kindly respond and suggest me

[@hemanthkumar1466]

Sir all are saying that to practice data sets so what exactly we should do with data sets plz reply

[@gargisingh9279]

Sir cant we compare bias and variance on the one random dataset? is it always comparison between two data set test error and conclude the variance ? or two dataset train error and compare the bias ?

[@VarunSingh-ds1hw]

hello sir i need your help ..that i want to get a data...basically a retail sales data that having promotional elements and different channels ... i am not able to find the data that exactly i need so can you please help me in this...

[@NguyenucNam_]

Hi, can you help me to answer problem that, I always at that we always want to low bias, so my purpose of the model only need to decrease bias? Right?

[@shadmanmartinpiyal4057]

Low bias and low variance both. If your model has only low bias and high variance that means the model is overfitting. If bias and variance both are high, it is underfitting. In layman's terms, target should be to have low error (training + test) depending on test selection which is done in k fold cross validation. You can refer to that video.

[@ruthvikrajam.v4303]

osm

[@192raghu]

Sir I always feel that your face look like Satya Nadella, Microsoft CEO.Did anyone say about you like this before sir?

[@codebasics]

Yes, you are probably a third person telling me this 🤓

[@192raghu]

@@codebasics sir.i have watched your git tutorials. Everything was done perfectly.but when try to push the code to github nothing happens sir.(I typed the command "git push" after commit).I have searched in Google to know the solution.some says setup proxy server (using git --config http proxy ..etc) . How to setup this proxy sir.(I have configure username and emai id on git bash) (Iam using mobile internet on my laptop).I wasting somuch time to know the solution for this issue.kindly help me sir

[@rafsangoni6979]

You shouldn't have smiled on the first pic xD

[@helloworld2740]

at First i am able to differentiate btween your faces I think its an face detect video😁😁😁😁

[@arepaconqueso4800]

Yayy

[@alwysrite]

variance and bias are analogous to accuracy and precision

[@dhruvsingh1111]

👹