This explanation blew my mind, sir. Washed away all wrong concepts. Thank you
@somanathking46942 ай бұрын
very clear and constructive!
@swapnanil70536 ай бұрын
Quite Informative, thank you sir
@tanayapanda16363 жыл бұрын
Unique topic
@UnfoldDataScience3 жыл бұрын
Thank you.
@souravbiswas68922 жыл бұрын
I have the issue understanding the DOF since last 5 years, but now I can say yes I got some idea. Thank you. 🙂
@UnfoldDataScience2 жыл бұрын
Glad it helped you
@balipavankalyan50088 ай бұрын
nice
@billinioreteachings3 жыл бұрын
Very good effort sir
@UnfoldDataScience3 жыл бұрын
Thanks a lot.
@unbiasedcrook91573 жыл бұрын
Sir can you please answer these two questions or share important links/references. Q1. Why it is desired for each feature to be normally distributed in case of regression problems? Q2. When we apply log-transformation on data to get normally distributed data, aren't we changing the data?
@UnfoldDataScience3 жыл бұрын
I may have to create a separate video. Will do.
@Julaiarvind3 жыл бұрын
Q1) residuals should be normally distributed as if the data is normally distributed we can make use of normal distribution and select the relevant features based on P values and considered model coeff. Q2)Applying transformation doesn't change the context of the data as the original value is represented by some other function.