期刊
NEUROCOMPUTING
卷 123, 期 -, 页码 424-435出版社
ELSEVIER
DOI: 10.1016/j.neucom.2013.08.004
关键词
Selective ensemble learning; Clustering; Dynamic selection; Circulating combination; Multi-label classification; Machine learning
资金
- Natural Science Foundation of China [61001013, 61102136, 61370010, 81101115]
- Natural Science Foundation of Fujian Province of China [2011J05158, 2011J01371, 2010J01351]
Selective ensemble is a learning paradigm that follows an overproduce and choose strategy, where a number of candidate classifiers are trained, and a set of several classifiers that are accurate and diverse are selected to solve a problem. In this paper, the hybrid approach called D3C is presented; this approach is a hybrid model of ensemble pruning that is based on k-means clustering and the framework of dynamic selection and circulating in combination with a sequential search method. Additionally, a multi-label D3C is derived from D3C through employing a problem transformation for multi-label classification. Empirical study shows that D3C exhibits competitive performance against other high-performance methods, and experiments in multi-label datasets verify the feasibility of multi-label D3C. (C) 2013 Elsevier B.V. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据