文献类型：期刊文献

英文题名：Geometric Structural Ensemble Learning for Imbalanced Problems

作者：Zhu, Zonghai[1,2];Wang, Zhe[1,2];Li, Dongdong[2];Zhu, Yujin[2];Du, Wenli[1]

机构：[1]East China Univ Sci & Technol, Minist Educ, Key Lab Adv Control & Optimizat Chem Proc, Shanghai 200237, Peoples R China;[2]East China Univ Sci & Technol, Dept Comp Sci & Engn, Shanghai 200237, Peoples R China

年份：2020

卷号：50

期号：4

起止页码：1617

外文期刊名：IEEE TRANSACTIONS ON CYBERNETICS

收录：;EI(收录号：20184606073779);WOS:【SCI-EXPANDED(收录号:WOS:000519727800023)】；

基金：This work was supported in part by the Natural Science Foundation of China under Grant 61672227, in part by the National Science Foundation of China for Distinguished Young Scholars under Grant 61725301, and in part by "Shuguang Program" through the Shanghai Education Development Foundation and the Shanghai Municipal Education Commission. This paper was recommended by Associate Editor Y. Jin.

语种：英文

外文关键词：Basic classifier; ensemble learning; geometric structure; imbalanced problems; machine learning; relaxation techniques

摘要：The classification on imbalanced data sets is a great challenge in machine learning. In this paper, a geometric structural ensemble (GSE) learning framework is proposed to address the issue. It is known that the traditional ensemble methods train and combine a series of basic classifiers according to various weights, which might lack the geometric meaning. Oppositely, the GSE partitions and eliminates redundant majority samples by generating hyper-sphere through the Euclidean metric and learns basic classifiers to enclose the minority samples, which achieves higher efficiency in the training process and seems easier to understand. In detail, the current weak classifier builds boundaries between the majority and the minority samples and removes the former. Then, the remaining samples are used to train the next. When the training process is done, all of the majority samples could be cleaned and the combination of all basic classifiers is obtained. To further improve the generalization, two relaxation techniques are proposed. Theoretically, the computational complexity of GSE could approach O(nd log(n(min)) log(n(maj))). The comprehensive experiments validate both the effectiveness and efficiency of GSE.