0

Linear SVM Regression: Primal Formula: Suppose we have a set of training data where xn is a multivariate set of N observations with observed response values yn.

To find the linear function

f(x)=x′β+b,

and ensure that it is as flat as possible.

What is the meaning of f(x) to be flat?

enter image description here

user weASD
  • 1
  • Where did you hear this? This is not a standard term. It would help if you could post a screenshot or an exact quote. – littleO Jun 04 '20 at 09:36
  • I added the screenshot. Please explain. Thank you. – user weASD Jun 04 '20 at 12:36
  • The phrasing is not very clear, but I think what it means is that if $\beta$ were exactly $0$ then $f$ would be constant (so its graph is perfectly "flat"). If $\beta$ is non-zero but small, then the graph of $f$ is in some sense close to being flat. We are trying to make $\beta$ as small as possible subject to the given constraints, so roughly speaking we are trying to make the graph of $f$ as "flat" as possible. – littleO Jun 04 '20 at 12:43

0 Answers0