R help - Jun 2005 - SVM parameters...

hi,

i am really sorry to ask this on the list, but i havent been able to
find anything on this topic.

i would like to know how the various parameters in the svm function
call in library e1071 work. all the literature that i was able to find
on the internet have been on the mathematics and derivation of
equations of the SVM or some very specific examples that relate to
biostatistics. i have been unable to find a concise description of how
these parameters affect the model.
so far i have been using a brute force method, running all possible
permutations and combinations , but this is taking enormous amounts of
time.

i would be grateful for any help that you could provide. 

thanks and regards ,
vivek.

Vivek,

I certainly would agree that every help page, including the one of
svm(), could be improved, but I think it is not _that_ deficient. In
particular, it tells you which parameters are used in the various
kernels available.

Have you read the corresponding article in R News (basically contained
as a vignette in the package)?

In addition, you could have a look at the documentation of libsvm, the
library that is interfaced by the svm()-function in e1071.

Best,
David

----------------

hi,

i am really sorry to ask this on the list, but i havent been able to
find anything on this topic.

i would like to know how the various parameters in the svm function
call in library e1071 work. all the literature that i was able to find
on the internet have been on the mathematics and derivation of
equations of the SVM or some very specific examples that relate to
biostatistics. i have been unable to find a concise description of how
these parameters affect the model.
so far i have been using a brute force method, running all possible
permutations and combinations , but this is taking enormous amounts of
time.

i would be grateful for any help that you could provide. 

thanks and regards ,
vivek.