A P

Hi Is there a reason why RVM classification has a parameter for setting the maximum number of solver iterations and RVM regression does not have this parameter? Could this parameter also be useful in regression? Thanks

OK, thanks very much for your help.

Thanks. Then epsilon between the lines in the picture at http://kernelsvm.tripod.com/ corresponds to the epsilon insensitivity parameter in dlib?

Hi I'm trying to understand the difference between epsilon and epsilon insensitivity in support vector regression, and also how epsilon in regression differs from the meaning of epsilon used in classification. Looking at this website (http://kernelsvm.tripod.com/), can you confirm that when you talk about epsilon in the documentation this corresponds to the width between the thick and dashed line in the first picture? The epsilon insensitivity would therefore be the measurement outside the lines,...