News for hackers. Stuff that matters.

A comment on Linear Regression.

Re:Linear Regression by wolfi1 on Friday June 19, 2020 @07:28AM (#23573294)
I would prefer an introduction via the Moore-Penrose Pseudo-Inverse. It's a lot easier, imho
[ Parent | on HN ]
Re:Linear Regression by xthestreams on Friday June 19, 2020 @09:32AM (#23574347)
Introducing linear regression with gradient descent can also be really confusing for a newcomer. Gradient descent solves a problem (nonlinear optimization) which does not exist here.
[ Parent | on HN ]
Re:Linear Regression by leovailati on Friday June 19, 2020 @11:23AM (#23575378)
I agree. What is the purpose of applying gradient descent to a linear regression problem? I mean, the pseudo inverse _is_ a closed form solution! Is it maybe more efficient for very large problems? I just don't get it.
[ Parent | on HN ]