[R] Problem in ksvm prediction

bharath bharath-bhushan.damodaran at irisa.fr
Thu Jun 11 16:05:26 CEST 2015


Dear all,

I am using ksvm in R for SVM classification. I trained the SVM classifier
with prob.model=TRUE. I have the two following outputs when I predicted.

1) label<-predict(model, TestData)
         I have classification accuracy about 93%, where as in another mode

2) Probvalues<-predict(model, TestData, type="probabilities")
    label<-max.col(Probvalues)

      I have the classification accuracy of 74%.

 I don't why there is a difference in two modes. It will be better if have
predict probability scores also for me.

Please let me know how to rectify the error or what could be the reason for
this difference?

Thanks in Advance

With Regards

Bharath Bhushan D



--
View this message in context: http://r.789695.n4.nabble.com/Problem-in-ksvm-prediction-tp4708504.html
Sent from the R help mailing list archive at Nabble.com.



More information about the R-help mailing list