Paper Title:
Ensemble Pruning for Data Dependant Learners
  Abstract

Ensemble pruning searches for a selective subset of members that performs as well as, or better than ensemble of all members. However, in the accuracy / diversity pruning framework, generalization ability of target ensemble is not considered, and moreover, there is not clear relationship between them. In this paper, we proof that ensemble formed by members of better generalization ability is also of better generalization ability. We adopt learning with both labeled and unlabeled data to improve generalization ability of member learners. A data dependant kernel determined by a set of unlabeled points is plugged in individual kernel learners to improve generalization ability, and ensemble pruning is launched as much previous work. The proposed method is suitable for both single-instance and multi-instance learning framework. Experimental results on 10 UCI data sets for single-instance learning and 4 data sets for multi-instance learning show that subensemble formed by the proposed method is effective.

  Info
Periodical
Chapter
Chapter 5: Information Technology
Edited by
Robin G. Qiu and Yongfeng Ju
Pages
522-527
DOI
10.4028/www.scientific.net/AMM.135-136.522
Citation
G. Zhang, S. H. Zhan, C. R. Wang, L. L. Cheng, "Ensemble Pruning for Data Dependant Learners", Applied Mechanics and Materials, Vols. 135-136, pp. 522-527, 2012
Online since
October 2011
Export
Price
$32.00
Share

In order to see related information, you need to Login.

In order to see related information, you need to Login.

Authors: Kai Li, Hong Tao Gao
Abstract:To improve the generalization performance for ensemble learning, a subgraph based selective classifier ensemble algorithm is presented....
261
Authors: Zheng Gang Gu, Kun Hong Liu
Abstract:Designing an evolutionary multiple classifier system (MCS) is a relatively new research area. In this paper, we propose a genetic algorithm...
2077
Authors: Da Ya Chen, Shang Ping Zhong
Chapter 6: Algorithm Design
Abstract:Universal steganalysis include feature extraction and steganalyzer design. Most universal steganalysis use Support Vector Machine (SVM) as...
1548
Authors: Qiong Wu, Ling Wei Wang, Jia Wu
Chapter 4: Data, Image and Signal Processing
Abstract:The characteristics of hyperspectral data with large number of bands, each bands have correlation, which has required a very high demand of...
508
Authors: Yang Li, Jian Lin Li, Song Jie Yue, Zhi Wang
Chapter 13: Network Engineering and Network Security
Abstract:In order to improve the detection accuracy of the intrusion detection, the proposed ensemble learning model is applied to the intrusion...
2376