Question: SVMs solve a classification problem by, in effect, transforming the feature space into another typically higher dimensional, sometimes infinitely dimensional space and then finding a
SVMs solve a classification problem by, in effect, transforming the feature space into another typically higher dimensional, sometimes infinitely dimensional space and then finding a (possibly soft) wide margin linear separation in that other space. What is the VC-dimension of linear classifications in an infinitely dimensional space? Is that a problem? If so, does the use of wide margin separations address that problem? Explain.
Step by Step Solution
There are 3 Steps involved in it
Get step-by-step solutions from verified subject matter experts
