DMCA
Dynamic Label Propagation for Semi-supervised Multi-class Multi-label Classification
| Citations: | 2 - 0 self |
BibTeX
@MISC{Wang_dynamiclabel,
author = {Bo Wang and Stanford Univeristy and Zhuowen Tu and John K. Tsotsos},
title = {Dynamic Label Propagation for Semi-supervised Multi-class Multi-label Classification},
year = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
In graph-based semi-supervised learning approaches, the classification rate is highly dependent on the size of the availabel labeled data, as well as the accuracy of the similarity measures. Here, we propose a semi-supervised multi-class/multi-label classification scheme, dynamic la-bel propagation (DLP), which performs transductive learn-ing through propagation in a dynamic process. Existing semi-supervised classification methods often have difficulty in dealing with multi-class/multi-label problems due to the lack in consideration of label correlation; our algorithm in-stead emphasizes dynamic metric fusion with label infor-mation. Significant improvement over the state-of-the-art methods is observed on benchmark datasets for both multi-class and multi-label tasks. 1.
Keyphrases
dynamic label propagation semi-supervised multi-class multi-label classification significant improvement semi-supervised multi-class multi-label classification scheme multi-class multi-label problem dynamic process semi-supervised classification method similarity measure state-of-the-art method multi-label task dynamic la-bel propagation label infor-mation transductive learn-ing benchmark datasets graph-based semi-supervised learning approach classification rate label correlation
