why my SVM don't have a support vector?

1 view (last 30 days)
JIE
JIE on 3 Sep 2014
Commented: Elisabetta Besana on 30 Mar 2017
Hi,all. I use SVM (matlab) to classify a dataset, but I come across some problem. when the training over, we will get a structure whose name is svmStruct, and it contains several fields. However, when I get my svmStruct, I find there is no support vectors and alpha, the result is as follows:
My code for training svm is:"svmStruct = svmtrain(Asvm_r', b_r,'Kernel_Function','polynomial','Polyorder',5);"
And I have 36 samples for label 1, and 28 samples for label -1. And when I use svmclassify with that "svmstrcut", it can work, and can classify the dataset.
I wonder why there are no support vectors and alpha in the structure "svmstruct" and how can the function svmclassify work with such a "svmstruct" ?
  1 Comment
Elisabetta Besana
Elisabetta Besana on 30 Mar 2017
I've got the same problem! Were you able to find an answer? If so can you explain it to me? Thank you so much

Sign in to comment.

Answers (0)

Tags

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!