Selective Ensemble Learning Method for Belief-Rule-Base Classification System Based on PAES
Traditional Belief-Rule-Based (BRB) ensemble learning methods integrate all of the trained sub-BRB systems to obtain better results than a single belief-rule-based system. However, as the number of BRB systems participating in ensemble learning increases, a large amount of redundant sub-BRB systems...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Tsinghua University Press
2019-12-01
|
| Series: | Big Data Mining and Analytics |
| Subjects: | |
| Online Access: | https://www.sciopen.com/article/10.26599/BDMA.2019.9020008 |
| _version_ | 1797990475534172160 |
|---|---|
| author | Wanling Liu Weikun Wu Yingming Wang Yanggeng Fu Yanqing Lin |
| author_facet | Wanling Liu Weikun Wu Yingming Wang Yanggeng Fu Yanqing Lin |
| author_sort | Wanling Liu |
| collection | DOAJ |
| description | Traditional Belief-Rule-Based (BRB) ensemble learning methods integrate all of the trained sub-BRB systems to obtain better results than a single belief-rule-based system. However, as the number of BRB systems participating in ensemble learning increases, a large amount of redundant sub-BRB systems are generated because of the diminishing difference between subsystems. This drastically decreases the prediction speed and increases the storage requirements for BRB systems. In order to solve these problems, this paper proposes BRBCS-PAES: a selective ensemble learning approach for BRB Classification Systems (BRBCS) based on Pareto-Archived Evolutionary Strategy (PAES) multi-objective optimization. This system employs the improved Bagging algorithm to train the base classifier. For the purpose of increasing the degree of difference in the integration of the base classifier, the training set is constructed by the repeated sampling of data. In the base classifier selection stage, the trained base classifier is binary coded, and the number of base classifiers participating in integration and generalization error of the base classifier is used as the objective function for multi-objective optimization. Finally, the elite retention strategy and the adaptive mesh algorithm are adopted to produce the PAES optimal solution set. Three experimental studies on classification problems are performed to verify the effectiveness of the proposed method. The comparison results demonstrate that the proposed method can effectively reduce the number of base classifiers participating in the integration and improve the accuracy of BRBCS. |
| first_indexed | 2024-04-11T08:35:59Z |
| format | Article |
| id | doaj.art-e2ac88eafc6d4db9baf25b9240573648 |
| institution | Directory Open Access Journal |
| issn | 2096-0654 |
| language | English |
| last_indexed | 2024-04-11T08:35:59Z |
| publishDate | 2019-12-01 |
| publisher | Tsinghua University Press |
| record_format | Article |
| series | Big Data Mining and Analytics |
| spelling | doaj.art-e2ac88eafc6d4db9baf25b92405736482022-12-22T04:34:19ZengTsinghua University PressBig Data Mining and Analytics2096-06542019-12-012430631810.26599/BDMA.2019.9020008Selective Ensemble Learning Method for Belief-Rule-Base Classification System Based on PAESWanling Liu0Weikun Wu1Yingming Wang2Yanggeng Fu3Yanqing Lin4<institution content-type="dept">School of Mathematics and Computer Science</institution>, <institution>Fuzhou University</institution>, <city>Fuzhou</city> <postal-code>350116</postal-code>, <country>China</country>.<institution content-type="dept">School of Mathematics and Computer Science</institution>, <institution>Fuzhou University</institution>, <city>Fuzhou</city> <postal-code>350116</postal-code>, <country>China</country>.<institution content-type="dept">Institute of Decision Sciences</institution>, <institution>Fuzhou University</institution>, <city>Fuzhou</city> <postal-code>350116</postal-code>, <country>China</country>.<institution content-type="dept">School of Mathematics and Computer Science</institution>, <institution>Fuzhou University</institution>, <city>Fuzhou</city> <postal-code>350116</postal-code>, <country>China</country>.<institution content-type="dept">School of Mathematics and Computer Science</institution>, <institution>Fuzhou University</institution>, <city>Fuzhou</city> <postal-code>350116</postal-code>, <country>China</country>.Traditional Belief-Rule-Based (BRB) ensemble learning methods integrate all of the trained sub-BRB systems to obtain better results than a single belief-rule-based system. However, as the number of BRB systems participating in ensemble learning increases, a large amount of redundant sub-BRB systems are generated because of the diminishing difference between subsystems. This drastically decreases the prediction speed and increases the storage requirements for BRB systems. In order to solve these problems, this paper proposes BRBCS-PAES: a selective ensemble learning approach for BRB Classification Systems (BRBCS) based on Pareto-Archived Evolutionary Strategy (PAES) multi-objective optimization. This system employs the improved Bagging algorithm to train the base classifier. For the purpose of increasing the degree of difference in the integration of the base classifier, the training set is constructed by the repeated sampling of data. In the base classifier selection stage, the trained base classifier is binary coded, and the number of base classifiers participating in integration and generalization error of the base classifier is used as the objective function for multi-objective optimization. Finally, the elite retention strategy and the adaptive mesh algorithm are adopted to produce the PAES optimal solution set. Three experimental studies on classification problems are performed to verify the effectiveness of the proposed method. The comparison results demonstrate that the proposed method can effectively reduce the number of base classifiers participating in the integration and improve the accuracy of BRBCS.https://www.sciopen.com/article/10.26599/BDMA.2019.9020008belief-rule-basepareto-archived evolutionary strategyselective ensembleclassification |
| spellingShingle | Wanling Liu Weikun Wu Yingming Wang Yanggeng Fu Yanqing Lin Selective Ensemble Learning Method for Belief-Rule-Base Classification System Based on PAES Big Data Mining and Analytics belief-rule-base pareto-archived evolutionary strategy selective ensemble classification |
| title | Selective Ensemble Learning Method for Belief-Rule-Base Classification System Based on PAES |
| title_full | Selective Ensemble Learning Method for Belief-Rule-Base Classification System Based on PAES |
| title_fullStr | Selective Ensemble Learning Method for Belief-Rule-Base Classification System Based on PAES |
| title_full_unstemmed | Selective Ensemble Learning Method for Belief-Rule-Base Classification System Based on PAES |
| title_short | Selective Ensemble Learning Method for Belief-Rule-Base Classification System Based on PAES |
| title_sort | selective ensemble learning method for belief rule base classification system based on paes |
| topic | belief-rule-base pareto-archived evolutionary strategy selective ensemble classification |
| url | https://www.sciopen.com/article/10.26599/BDMA.2019.9020008 |
| work_keys_str_mv | AT wanlingliu selectiveensemblelearningmethodforbeliefrulebaseclassificationsystembasedonpaes AT weikunwu selectiveensemblelearningmethodforbeliefrulebaseclassificationsystembasedonpaes AT yingmingwang selectiveensemblelearningmethodforbeliefrulebaseclassificationsystembasedonpaes AT yanggengfu selectiveensemblelearningmethodforbeliefrulebaseclassificationsystembasedonpaes AT yanqinglin selectiveensemblelearningmethodforbeliefrulebaseclassificationsystembasedonpaes |