consider switching to xy interface for `kernlab::ksvm()`

[simonpcouch]

Related to #1114.

This SO post runs into troubles with the formula interface of kernlab::ksvm() which could be resolved by using the matrix interface of the kernlab function.

We should explore whether there are any potential downsides to switching to kernlab's xy interface and confirm that it resolves the memory protection issues seen in that post.

[clabornd]

Cant speak to any potential downsides, but fwiw changing svm_poly to use the matrix interface allowed me to run examples that were previously giving me the protection stack overflow error (20K+ features) and have consistent results with the formula interface in examples where both run without error.

Also wondering if directly modifying get_model_env()$svm_poly_fit is a good temporary workaround?