IIIT Hyderabad Publications |
|||||||||
|
Semi-Supervised Clustering by Selecting Informative ConstraintsAuthors: Vidyadhar Rao, C V Jawahar Conference: PReMI 2013 Date: 2013-12-10 Report no: IIIT/TR/2013/113 AbstractTraditional clustering algorithms use a predened metric and no supervision in identifying the partition. Existing semi-supervised clus- tering approaches either learn a metric from randomly chosen constraints or actively select informative constraints using a generic distance measure like Euclidean norm. We tackle the problem of identifying constraints that are informative to learn appropriate metric for semi-supervised clus- tering. We propose an approach to simultaneously nd out appropriate constraints and learn a metric to boost the clustering performance. We evaluate clustering quality of our approach using the learned metric on the MNIST handwritten digits, Caltech-256 and MSRC2 object image datasets. Our results on these datasets have signicant improvements over the baseline methods like MPCK-MEANS. Full paper: pdf Centre for Visual Information Technology |
||||||||
Copyright © 2009 - IIIT Hyderabad. All Rights Reserved. |