A Feature Selection Method Based on Competition Winners Mechanism

DOI

10.2991/ipemec-15.2015.118How to use a DOI?

Keywords

Text Categorization;Feature Selection;Information Gain

Abstract

Feature selection plays an important role in the field of text categorization. The traditional feature selection methods such as information gain(IG), the weight of evidence for text(WET) , feature selection and so on are commonly applied in text categorization. However, the traditional feature selection methods are based on local features, there are many low information redundancies features was selected. In this paper a new feature selection algorithm called CWFS based on competition winners feature selection is proposed to solve the time consuming issue of classification as well as low accuracy. By removing uncorrelated or redundant features, the filter model only applies the top dozens of all features, thus the filter model can finish the classification with less time. Our concern is to reduce the dimension of the feature space. The classifiers with Naive Bayes and support vector machine have been used to run our experiment on TREC sets. The experimental results show that CWFS method can highly improve the quality of classification.

Copyright

© 2015, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Chengyan Li
AU  - Xiaodong Wang
PY  - 2015/05
DA  - 2015/05
TI  - A Feature Selection Method Based on Competition Winners Mechanism
BT  - Proceedings of the 2015 International Power, Electronics and Materials Engineering Conference
PB  - Atlantis Press
SP  - 638
EP  - 643
SN  - 2352-5401
UR  - https://doi.org/10.2991/ipemec-15.2015.118
DO  - 10.2991/ipemec-15.2015.118
ID  - Li2015/05
ER  -