A Developed Feature Selection Method for Classification Based on United Information Gain

被引：0

作者：

Niu, Kun ^{[1
]}

Jiao, Haizhen ^{[1
]}

Gao, Zhipeng ^{[2
]}

Jia, Guannan ^{[1
]}

Yang, Guangyu ^{[1
]}

Cheng, Cheng ^{[3
]}

机构：

[1] Beijing Univ Posts & Telecommun, Sch Software Engn, 10 Xitucheng Rd, Beijing 100876, Peoples R China

[2] Beijing Univ Posts & Telecommun, Inst Network Technol, 10 Xitucheng Rd, Beijing 100876, Peoples R China

[3] Beijing Univ Posts & Telecommun, Sch Informat & Commun Engn, 10 Xitucheng Rd, Beijing 100876, Peoples R China

来源：

2017 IEEE SMARTWORLD, UBIQUITOUS INTELLIGENCE & COMPUTING, ADVANCED & TRUSTED COMPUTED, SCALABLE COMPUTING & COMMUNICATIONS, CLOUD & BIG DATA COMPUTING, INTERNET OF PEOPLE AND SMART CITY INNOVATION (SMARTWORLD/SCALCOM/UIC/ATC/CBDCOM/IOP/SCI) | 2017年

基金：

中国国家自然科学基金;

关键词：

Classification; Feature Selection; United Information Gain; MUTUAL INFORMATION;

D O I：

暂无

中图分类号：

TP301 [理论、方法];

学科分类号：

081202 ;

摘要：

In big data times, effective feature selection methods become more and more important because of high dimensional character. In this paper, we propose a developed feature selection method UIG, what effectively improves the accuracy of classifiers. Firstly, all the attributes are systemized pairwise to calculate their united information gain to form a United Information Gain Matrix. Secondly, the United Information Gain Matrix is discretized by a predefined parameter. Finally, UIG searches the biggest full-1 sub-matrix to find the maximum valid subset of attributes. Then classifier can be built by the selected attributes. Experiments on several public data sets show that the proposed method is effective and robust for different situations.

引用

页数：4

共 50 条

[31] Mutual information-based feature selection for multilabel classification
Doquire, Gauthier
Verleysen, Michel
[J]. NEUROCOMPUTING, 2013, 122 : 148 - 155
[32] Feature ranking based on information gain for large classification problems with MapReduce
Zdravevski, Eftim
Lameski, Petre
Kulakov, Andrea
Jakimovski, Boro
Filiposka, Sonja
Trajanov, Dimitar
[J]. 2015 IEEE TRUSTCOM/BIGDATASE/ISPA, VOL 2, 2015, : 186 - 191
[33] Mutual information based input feature selection for classification problems
Cang, Shuang
Yu, Hongnian
[J]. DECISION SUPPORT SYSTEMS, 2012, 54 (01) : 691 - 698
[34] A Feature Selection Approach Based on Information Theory for Classification Tasks
Jesus, Jhoseph
Canuto, Anne
Araujo, Daniel
[J]. ARTIFICIAL NEURAL NETWORKS AND MACHINE LEARNING, PT II, 2017, 10614 : 359 - 367
[35] Binary Differential Evolution based Feature Selection Method with Mutual Information for Imbalanced Classification Problems
Ghosh, Arka
Xue, Bing
Zhang, Mengjie
[J]. 2021 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION (CEC 2021), 2021, : 794 - 801
[36] A Novel Feature Selection Method Based on Category Information Analysis for Class Prejudging in Text Classification
Wang, Qiang
Guan, Yi
Wang, XiaoLong
Xu, Zhiming
[J]. INTERNATIONAL JOURNAL OF COMPUTER SCIENCE AND NETWORK SECURITY, 2006, 6 (1A): : 113 - 119
[37] A new feature selection method for handling redundant information in text classification
Wang, You-wei
Feng, Li-zhou
[J]. FRONTIERS OF INFORMATION TECHNOLOGY & ELECTRONIC ENGINEERING, 2018, 19 (02) : 221 - 234
[38] A new feature selection method for handling redundant information in text classification
You-wei Wang
Li-zhou Feng
[J]. Frontiers of Information Technology & Electronic Engineering, 2018, 19 : 221 - 234
[39] An Unsupervised Feature Selection Method Based on Information Entropy
Wang, Xiaohong
He, Yidi
Wang, Lizhi
Wang, Zhongxing
[J]. 2018 3RD INTERNATIONAL CONFERENCE ON SYSTEM RELIABILITY AND SAFETY (ICSRS), 2018, : 35 - 39
[40] A new feature selection method based on symmetrical uncertainty and interaction gain
Lin, Xiaohui
Li, Chao
Ren, Weijie
Luo, Xiao
Qi, Yanpeng
[J]. COMPUTATIONAL BIOLOGY AND CHEMISTRY, 2019, 83

← 1 2 3 4 5 →