Partial Classifier Chains with Feature Selection by Exploiting Label Correlation in Multi-Label Classification
Multi-label classification (MLC) is a supervised learning problem where an object is naturally associated with multiple concepts because it can be described from various dimensions. How to exploit the resulting label correlations is the key issue in MLC problems. The classifier chain (CC) is a well-...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-10-01
|
| Series: | Entropy |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1099-4300/22/10/1143 |
| _version_ | 1827704623477030912 |
|---|---|
| author | Zhenwu Wang Tielin Wang Benting Wan Mengjie Han |
| author_facet | Zhenwu Wang Tielin Wang Benting Wan Mengjie Han |
| author_sort | Zhenwu Wang |
| collection | DOAJ |
| description | Multi-label classification (MLC) is a supervised learning problem where an object is naturally associated with multiple concepts because it can be described from various dimensions. How to exploit the resulting label correlations is the key issue in MLC problems. The classifier chain (CC) is a well-known MLC approach that can learn complex coupling relationships between labels. CC suffers from two obvious drawbacks: (1) label ordering is decided at random although it usually has a strong effect on predictive performance; (2) all the labels are inserted into the chain, although some of them may carry irrelevant information that discriminates against the others. In this work, we propose a partial classifier chain method with feature selection (PCC-FS) that exploits the label correlation between label and feature spaces and thus solves the two previously mentioned problems simultaneously. In the PCC-FS algorithm, feature selection is performed by learning the covariance between feature set and label set, thus eliminating the irrelevant features that can diminish classification performance. Couplings in the label set are extracted, and the coupled labels of each label are inserted simultaneously into the chain structure to execute the training and prediction activities. The experimental results from five metrics demonstrate that, in comparison to eight state-of-the-art MLC algorithms, the proposed method is a significant improvement on existing multi-label classification. |
| first_indexed | 2024-03-10T15:44:28Z |
| format | Article |
| id | doaj.art-8798793360b6484bb6400ea288de411a |
| institution | Directory Open Access Journal |
| issn | 1099-4300 |
| language | English |
| last_indexed | 2024-03-10T15:44:28Z |
| publishDate | 2020-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Entropy |
| spelling | doaj.art-8798793360b6484bb6400ea288de411a2023-11-20T16:36:00ZengMDPI AGEntropy1099-43002020-10-012210114310.3390/e22101143Partial Classifier Chains with Feature Selection by Exploiting Label Correlation in Multi-Label ClassificationZhenwu Wang0Tielin Wang1Benting Wan2Mengjie Han3Department of Computer Science and Technology, China University of Mining and Technology, Beijing 100083, ChinaDepartment of Computer Science and Technology, China University of Mining and Technology, Beijing 100083, ChinaSchool of Software and IoT Engineering, Jiangxi University of Finance & Economics, Nanchang 330013, ChinaSchool of Technology and Business Studies, Dalarna University, 79188 Falun, SwedenMulti-label classification (MLC) is a supervised learning problem where an object is naturally associated with multiple concepts because it can be described from various dimensions. How to exploit the resulting label correlations is the key issue in MLC problems. The classifier chain (CC) is a well-known MLC approach that can learn complex coupling relationships between labels. CC suffers from two obvious drawbacks: (1) label ordering is decided at random although it usually has a strong effect on predictive performance; (2) all the labels are inserted into the chain, although some of them may carry irrelevant information that discriminates against the others. In this work, we propose a partial classifier chain method with feature selection (PCC-FS) that exploits the label correlation between label and feature spaces and thus solves the two previously mentioned problems simultaneously. In the PCC-FS algorithm, feature selection is performed by learning the covariance between feature set and label set, thus eliminating the irrelevant features that can diminish classification performance. Couplings in the label set are extracted, and the coupled labels of each label are inserted simultaneously into the chain structure to execute the training and prediction activities. The experimental results from five metrics demonstrate that, in comparison to eight state-of-the-art MLC algorithms, the proposed method is a significant improvement on existing multi-label classification.https://www.mdpi.com/1099-4300/22/10/1143multi-label classificationclassifier chainslabel correlationfeature selection |
| spellingShingle | Zhenwu Wang Tielin Wang Benting Wan Mengjie Han Partial Classifier Chains with Feature Selection by Exploiting Label Correlation in Multi-Label Classification Entropy multi-label classification classifier chains label correlation feature selection |
| title | Partial Classifier Chains with Feature Selection by Exploiting Label Correlation in Multi-Label Classification |
| title_full | Partial Classifier Chains with Feature Selection by Exploiting Label Correlation in Multi-Label Classification |
| title_fullStr | Partial Classifier Chains with Feature Selection by Exploiting Label Correlation in Multi-Label Classification |
| title_full_unstemmed | Partial Classifier Chains with Feature Selection by Exploiting Label Correlation in Multi-Label Classification |
| title_short | Partial Classifier Chains with Feature Selection by Exploiting Label Correlation in Multi-Label Classification |
| title_sort | partial classifier chains with feature selection by exploiting label correlation in multi label classification |
| topic | multi-label classification classifier chains label correlation feature selection |
| url | https://www.mdpi.com/1099-4300/22/10/1143 |
| work_keys_str_mv | AT zhenwuwang partialclassifierchainswithfeatureselectionbyexploitinglabelcorrelationinmultilabelclassification AT tielinwang partialclassifierchainswithfeatureselectionbyexploitinglabelcorrelationinmultilabelclassification AT bentingwan partialclassifierchainswithfeatureselectionbyexploitinglabelcorrelationinmultilabelclassification AT mengjiehan partialclassifierchainswithfeatureselectionbyexploitinglabelcorrelationinmultilabelclassification |