期刊
PROCEEDINGS OF THE 16TH INTERNATIONAL CONFERENCE ON HYBRID INTELLIGENT SYSTEMS (HIS 2016)
卷 552, 期 -, 页码 244-255出版社
SPRINGER INTERNATIONAL PUBLISHING AG
DOI: 10.1007/978-3-319-52941-7_25
关键词
Multi-label learning; Imbalanced data; Classification; Rank SVM
资金
- National Natural Science Foundation of China [61502091]
- Fundamental Research Funds for the Central Universities [N140403004]
- Postdoctoral Science Foundation of China [2015M570254]
Multi-label data classification has become an important and active research topic, where the classification algorithm is required to deal with prediction of sets of label indicators for instances simultaneously. Label powerset (LP) method reduces the multi-label classification problem to a single-label multi-class classification problem by treating each distinct combination of labels. However, the predictive performance of LP is challenged with imbalanced distribution among the labelsets, deteriorating the performance of traditional classifiers. In this paper, we study the problem of multi-label imbalanced data classification and propose a novel solution, called CSRankSVM (Cost sensitive Ranking Support Vector Machine), which assigns a different mis-classification cost for each labelset to effectively tackle the problem of imbalance for Multi-label data. Empirical studies on popular benchmark datasets with various imbalance ratios of labelsets demonstrate that the proposed CSRankSVM approach can effectively boost classification performances in multi-label datasets.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据