What is Multicollinearity? Extensive video + simulation!

You are the best!!!

Man you prepared me for a killer interview, thanks

It's Always Sunny in Philadelphia: the gang start a law firm

Hi Justin, really a very nice and insightful video but when we have a multiple linear regression with hundreds of features with heteroscedasticity, how we figure out which ones to log and which not ?

you are too good bro.... wish that every university to have a professor like you ..

I don't attend my stats lectures no more, I just watch Zed video and pass my tests . Thanks for the uid bro!! <3 from S. Africa

I dont understand any more😢😢

Thank you so much!

How can we not be certain of the coefficients and still have predictive power? where is the prediction power is coming from if not from the coefficients themselves.

I never thought I'd stumble upon a statistics video with not only a Sufjan Stevens vinyl in the background but also with It's Always Sunny references, immediately subscribed.

Why is vif = 1/(1-r2), if simply higher r2 mean higher vif. why define vif at all ?

Just discovered this channel !!! Am having fun

Thanks!

Thank you!

Genuinely, thank you

thank you

I really appreciate the intuitive explanation! Very valuable! Thanks a lot!

Thank you for going into the small details that are usually not explained well in schools. I appreciate your efforts. And I really enjoy your accent - charming. ;)

Janet looks skeptical rather than cynical.

Great job! Thank you!

underrated channel...

Thanks Justin, always a pleasure and transformative to watch your stats videos.

Best explanation of multicollinearity I've come across

Thanks for providing such clarity in so simple words. Love frm 🇮🇳

Great stuff

Very nice set of lectures. Thank you for your demo at the end!

So basically speaking when "X" variables have the same unit, one of them can be removed, either it is distance, time or cucumbers?

Exceptional explanation 👍

Thank you! Thank you! Thank you!

u r fuking good love u and videos

This is a complete break down of the concept of multicollinearity. Thanks so much man. Now about the 3rd remedy for multicollinearity you provided in the video, i was studying the effect of education on health status and to measure education i used gross enrollment at primary, gross enrollment at secondary and gross enrolment at tertiary levels but after running my regression i noticed the 3 variables were correlated. would i be right to combine the 3 variables by looking at the average to get what i termed "gross enrolment rate Total"?

I'm in a class and not understanding the way they are trying to teach me these concept. I really appreciate you work and the way you explain things. Not sure why I'm taking a $2,000 class when I can learn more for the price of a "like" on your videos.

Excellent teaching skills, really nice presentation, Thanks man

THANK YOU SO MUCH

Brilliant Video. you are just great. Very well explained.

Your videos are amazziiinggggg!!!!!!!

This was great. I've been teaching myself linear algebra over the past 6 months and it's cools to see a real world example of ill-conditioned matrices and, in the final example, of a coefficient matrix who's columns don't form a basis. The real world context really helps cement those more abstract concepts.

Hi, great video. What is the little epsilon at the end of the regression equation. I assume beta0 already is the intercept?

great explanation!
Only watch the example at the beginning, I knew multicollinearty is!!! Thank you

The way how you explained multicollinearity is awesome. Thank you for clearing my doubts.

can we disregard multi collinearity when developing a prediction model? My model is getting quite complex (with some terms with higher order to satisfy linearity on logit scale [fractional polynomials]) and its getting tricky to deal with this issue.

Why did I go to school!!!!!

Thank you much mate. :)

Very useful. Thank you!

great content . lean a lot. thankx man

This video is perfect! Thank you :D

can we use VIF when assessing multicollinearity (association) between categorical X variables (multinomial)? thank you!

Indeed great video. Thank you
I have a question Pearson's correlation between variables are statistically significant however in linear multiple regression it's now...why?
Something signs positive or negative are also not consistent with each other in correlation and regression...why?