From my understanding a large value of the C parameter in SVM decreases the number of misclassified instances and narrows the margin. My question is - considering that SVM with linear kernel is used, with soft margin - does increasing the value of C always guarantees that the margin is not getting bigger or there is a chance that it still can?
$\begingroup$
$\endgroup$
New contributor
BeyondSky is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.
Add a comment
|