Combining One-Class Classifiers via Meta-Learning

Details Combining One-Class Classifiers via Meta-Learning Combining One-Class Classifiers via Meta-Learning

2011-12-22

arxiv.org/abs/1112.5246v1

Computer Science

Learning

Related to both Ensemble learning and one-class learning. Length of 6 pages. This document is the smaller version of a journal

Scientific paper

We examine various methods for combining the output of one-class models. In particular, we show that simple meta-learning based ensemble achieves better result than weighting methods. Furthermore we propose a new one-class ensemble scheme, called TUPSO that uses meta-learning for combining multiple one-class classifiers. We also present a new one-class classification performance measures to weigh the base-classifiers, a process that proved helpful for increasing the classification performance of the induced ensemble. Our experimental study shows that the proposed method significantly outperforms exiting methods.

Affiliated with

Also associated with

No associations

Rating

Combining One-Class Classifiers via Meta-Learning does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.

If you have personal experience with Combining One-Class Classifiers via Meta-Learning, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Combining One-Class Classifiers via Meta-Learning will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFWR-SCP-O-305987