You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have few questions about the MSVR parameters, to which I could not find answers in papers or in source code.
The command line parameter param is transformed in main.cpp (line 66) as follows: double param = mult*mean_all(pdist(Xtest)); where mult is the value received from command line. param is then used in calculation of the kernel matrix in k_mat. Why is this transformation performed, i.e., why mult is not passed directly to the k_mat?
Is there any recommended range for optimization of the epsilon parameter?
The text was updated successfully, but these errors were encountered:
Hi!
I have few questions about the MSVR parameters, to which I could not find answers in papers or in source code.
The command line parameter param is transformed in main.cpp (line 66) as follows:
double param = mult*mean_all(pdist(Xtest));
wheremult
is the value received from command line.param
is then used in calculation of the kernel matrix ink_mat
. Why is this transformation performed, i.e., whymult
is not passed directly to thek_mat
?Is there any recommended range for optimization of the epsilon parameter?
The text was updated successfully, but these errors were encountered: