Multiple Linear Regression

Hey @Yash_Barve,
Welcome to the community. Just to add to @shanup’s explanation, I would like to mention that Prof. Andrew in the lecture videos explicitly mentioned about using simultaneous updates. The primary reason as to why simultaneous updates are preferred over the non-simultaneous ones is computational efficiency. You can check out this great discussion which goes one step ahead. It keeps the computational efficiency aside just for fun, and then discusses which of the 2 approaches might be more suitable. I hope this helps.

Regards,
Elemento

1 Like