计算机科学与探索2017，Vol.11Issue(4)：619-632,14.DOI:10.3778/j.issn.1673-9418.1603094
类不平衡模糊加权极限学习机算法研究
Research on Class Imbalance Fuzzy Weighted Extreme Learning Machine Algorithm
摘要
Abstract
Firstly,this paper analyzes the reason that the performance of extreme leaming machine (ELM) is destroyed by imbalanced instance distribution in theory.Then,based on the same theoretical framework,this paper discusses the effectiveness and inherent shortcomings of the weighted extreme learning machine (WELM).Nextly,profiting from the idea of fuzzy set,this paper proposes four fuzzy weighted extreme leaming machine (FWELM)algorithms to deal with class imbalance problem.Finally,this paper verifies the effectiveness and feasibility of these four FWELM algorithms by the experiments constructing on 20 baseline binary-class imbalanced data sets.The experimental results indicate that the proposed algorithms can often acquire better classification performance than WELM algorithm and several traditional class imbalance leaming algorithms in the context of ELM.In addition,in contrast with fuzzy support vector machine for class imbalance learning (FSVM-CIL) series algorithms,the proposed algorithms can produce the comparable classification performance,but always consume less training time.关键词
极限学习机/类不平衡学习/模糊加权/先验分布信息Key words
extreme learning machine/class imbalance learning/fuzzy weighting/prior distribution information分类
信息技术与安全科学引用本文复制引用
于化龙,祁云嵩,杨习贝,左欣..类不平衡模糊加权极限学习机算法研究[J].计算机科学与探索,2017,11(4):619-632,14.基金项目
The National Natural Science Foundation of China under Grant Nos.61305058,61471182,61572242(国家自然科学基金) （国家自然科学基金）
the Natural Science Foundation of Jiangsu Province under Grant Nos.BK20130471,BK20150470(江苏省自然科学基金) （江苏省自然科学基金）
the Postdoctoral Science Foundation of China under Grant Nos.2013M540404,2015T80481(中国博士后科学基金) （中国博士后科学基金）
the Postdoctoral Research Funds of Jiangsu Province under Grant No.1401037B(江苏省博士后基金). （江苏省博士后基金）
