Building models with SAS Enterprise Miner, SAS Factory Miner, SAS Visual Data Mining and Machine Learning or just with programming

Is there any way to implement one-class SVM for anomaly detection in EM 14.1?

Reply
Occasional Contributor
Posts: 16

Is there any way to implement one-class SVM for anomaly detection in EM 14.1?

Hello,

 

I am analysing a dataset with highly unbalanced data, and would like to try out different approaches, one of which is also anomaly detection approach. Does somebody know if there is a way to use one class SVM in EM, like an existing custom SAS code or some tweaking of the existing SVM node?

 

Thank you Smiley Happy

 

P.S. article about one class SVM - http://rvlasveld.github.io/blog/2013/07/12/introduction-to-one-class-support-vector-machines/

SAS Super FREQ
Posts: 306

Re: Is there any way to implement one-class SVM for anomaly detection in EM 14.1?

Not yet, but keep your eyes peeled!  In a release in the near-future, a Viya procedure is coming for implementing the Support Vector Data Description method mentioned in your link.

Occasional Contributor
Posts: 16

Re: Is there any way to implement one-class SVM for anomaly detection in EM 14.1?

Posted in reply to WendyCzika

Thanks a lot for letting me know, I'm looking forward to it Smiley Happy 

Ask a Question
Discussion stats
  • 2 replies
  • 304 views
  • 0 likes
  • 2 in conversation