08-15-2017 02:01 PM
The short answers are as follows:
* Interpretation - I am not aware of any broadly accepted ways to interpret an SVM model.
* Implementation - You can use the HP SVM node -- which is found on the HPDM tab in SAS Enterprise Miner -- to fit an SVM model, or you can use the HPSVM procedure.
* Documentation - You can find help for the HP SVM node once you open SAS Enterprise Miner by clicking on Help --> Contents and then navigating in the panel on the left to
HP SVM Node
Then look in the panel on the right in order to the available help topics. For the HPSVM procedure, open your web browser to
and click on the link to
SAS Enterprise Miner 14.2 High-Performance Procedures
Here is an excerpt from the SAS Enterprise Miner help utility:
The SAS Enterprise Miner HP SVM Node uses PROC HPSVM and PROC HPSVMSCORE. The HP SVM Node supports only binary classification problems, including polynomial, radial basis function, and sigmoid nonlinear kernels. The HP SVM Node does not perform multi-class problems or support vector regression.
I hope this helps!