Ridge regression - what is k=0?

Question

I'm getting to know ridge regression and what to check my understanding quickly.

I understand that k is the shrinkage parameter. If I'm reading off coefficients where k=0, is that equivalent to an OLS linear regression?

Or have I got totally the wrong end of the stick?

Simon.

score 2 · Accepted Answer · answered Oct 15 '14 at 22:13

2

Assuming k is the coefficient in front of the regularization term, then yes having k=0 is the same as OLS linear regression.

answered Oct 15 '14 at 22:13

Aaron

1

Excellent - yes, I believe that we are describing the same parameter using different terms (though please anybody correct me if you think that I am wrong). Thank you so much, Aaron. – SimonsSchus Oct 15 '14 at 22:27
Of course. when you have k equal to zero, you will return to the problem of minimizing $||y - X\beta||^2$ w.r.t $\beta$ . – Christina May 12 '15 at 13:50

1 Answers1