A novel oversampling technique based on the manifold distance for class imbalance learning
文献类型:期刊论文
作者 | Guo YN(郭一楠)2; Jiao BT(焦博韬)2; Yang LK(杨凌凯)2; Cheng J(程健)4; Yang SX(杨圣祥)1; Tang FZ(唐凤珍)3 |
刊名 | INTERNATIONAL JOURNAL OF BIO-INSPIRED COMPUTATION |
出版日期 | 2021 |
卷号 | 18期号:3页码:131-142 |
ISSN号 | 1758-0366 |
关键词 | class imbalance learning oversampling manifold learning overlapping small disjunction |
产权排序 | 4 |
英文摘要 | Oversampling is a popular problem-solver for class imbalance learning by generating more minority samples to balance the dataset size of different classes. However, resampling in original space is ineffective for the imbalance datasets with class overlapping or small disjunction. Based on this, a novel oversampling technique based on manifold distance is proposed, in which a new minority sample is produced in terms of the distances among neighbours in manifold space, rather than the Euclidean distance among them. After mapping the original data to its manifold structure, the overlapped majority and minority samples will lie in areas easily being partitioned. In addition, the new samples are generated based on the neighbours locating nearby in manifold space, avoiding the adverse effect of the disjoint minority classes. Following that, an adaptive adjustment method is presented to determine the number of the newly generated minority samples according to the distribution density of the matched-pair data. The experimental results on 48 imbalanced datasets indicate that the proposed oversampling technique has the better classification accuracy. |
WOS关键词 | OPTIMIZATION ; ENSEMBLE |
资助项目 | National Natural Science Foundation of China[61973305] ; National Natural Science Foundation of China[61573361] ; National Natural Science Foundation of China[61803369] ; Natural Science Foundation of Liaoning Province for the State Key Laboratory of Robotics[2020-KF-22-02] ; State Key Laboratory of Robotics[2019-O12] |
WOS研究方向 | Computer Science |
语种 | 英语 |
WOS记录号 | WOS:000724400800001 |
资助机构 | National Natural Science Foundation of ChinaNational Natural Science Foundation of China (NSFC) [61973305, 61573361, 61803369] ; Natural Science Foundation of Liaoning Province for the State Key Laboratory of Robotics [2020-KF-22-02] ; State Key Laboratory of Robotics [2019-O12] |
源URL | [http://ir.sia.cn/handle/173321/30098] |
专题 | 沈阳自动化研究所_机器人学研究室 |
通讯作者 | Cheng J(程健) |
作者单位 | 1.De Montfort University, Leicester LE1 9BH, UK 2.School of Information and Control Engineering, China University of Mining and Technology, Xuzhou, China 3.Shenyang Institute of Automation, Shenyang, China 4.China Coal Research Institute, Beijing 100013, China |
推荐引用方式 GB/T 7714 | Guo YN,Jiao BT,Yang LK,et al. A novel oversampling technique based on the manifold distance for class imbalance learning[J]. INTERNATIONAL JOURNAL OF BIO-INSPIRED COMPUTATION,2021,18(3):131-142. |
APA | Guo YN,Jiao BT,Yang LK,Cheng J,Yang SX,&Tang FZ.(2021).A novel oversampling technique based on the manifold distance for class imbalance learning.INTERNATIONAL JOURNAL OF BIO-INSPIRED COMPUTATION,18(3),131-142. |
MLA | Guo YN,et al."A novel oversampling technique based on the manifold distance for class imbalance learning".INTERNATIONAL JOURNAL OF BIO-INSPIRED COMPUTATION 18.3(2021):131-142. |
入库方式: OAI收割
来源:沈阳自动化研究所
其他版本
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。