Topic: Complex Predictive Models

Which statements about optimizing the parameters of the Support Vector Machines operator are correct? $($Select ANY correct answer$)$

A$.$

Optimizing SVMs is difficult because there is an unbounded parameter space, and sometimes small changes in parameters can make a big difference.

B$.$

There is often more than one way to get near$-$optimal results.

C$.$

An automated approach $($e$.$g$.$ using Optimize Parameters $($Evolutionary$))$ can be helpful.

D$.$

It is important to understand the details of each kernel type in order to select the correct kernel.