Hi all,
Could someone please recommend software for svm regression with around 1
million observations and around 200 explanatory variables? A matlab
implementation would be preferred.
I was disappointed that no-one responded to a previous request for
theoretical results concerning svm performance in speficially time-series
applications (only one person responded saying they were interested in any
answers I got). Is this because such results do not exist?
Thanks,
Spyros
|