Ask Your Question
0

SVM: call train() multiple times?

asked 2016-06-10 06:20:38 -0500

RoughHough gravatar image

Hello there!

I'm using openCV's SVM for classification problems (obviously) and I was just wondering how to train it properly. For now, my program reads a number of files with serialised feature vectors and calls CvSVM::train(...) in a for-loop. Is this the correct way to do or do I have to merge all feature vectors into a big one and pass this one then? It's a single SVM object all the time btw.

Thanks and have a great day,

RoughHough

edit retag flag offensive close merge delete

1 answer

Sort by ยป oldest newest most voted
2

answered 2016-06-10 18:46:36 -0500

berak gravatar image

For now, my program reads a number of files with serialised feature vectors and calls CvSVM::train(...) in a for-loop. Is this the correct way to do or do I have to merge all feature vectors into a big one

no, the former is wrong, and the latter is the correct way.

edit flag offensive delete link more
Login/Signup to Answer

Question Tools

1 follower

Stats

Asked: 2016-06-10 06:20:38 -0500

Seen: 174 times

Last updated: Jun 10 '16