Binary Imbalanced Data Classification Based on Modified D2GAN Oversampling and Classifier Fusion
Binary imbalance problem refers to such a classification scenario where one class contains a large number of samples while another class contains only a few samples. When traditional classifiers face with imbalanced datasets, they usually bias towards majority class resulting in poor classification...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2020-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9195865/ |
| _version_ | 1818936251423653888 |
|---|---|
| author | Junhai Zhai Jiaxing Qi Sufang Zhang |
| author_facet | Junhai Zhai Jiaxing Qi Sufang Zhang |
| author_sort | Junhai Zhai |
| collection | DOAJ |
| description | Binary imbalance problem refers to such a classification scenario where one class contains a large number of samples while another class contains only a few samples. When traditional classifiers face with imbalanced datasets, they usually bias towards majority class resulting in poor classification performance. Oversampling is an effective method to address this problem, yet how to conduct diversity oversampling is a challenge. In this article, we proposed a diversity oversampling method based on a modified D2GAN model, and on the basis of diversity oversampling, we also proposed a binary imbalanced data classification approach based on classifier fusion by fuzzy integral. Extensive experiments are conducted on 8 data sets to compare the proposed methods with 7 state-of-the-art methods on 5 aspects: MMD-score, Silhouette-score, F-measure, G-means, and AUC-area. The 7 methods include 4 SMOTE related approaches and 3 GAN related approaches. The experimental results demonstrate that the proposed methods are more effective and efficient than the compared approaches. |
| first_indexed | 2024-12-20T05:33:06Z |
| format | Article |
| id | doaj.art-3248186172f54b829d379f64d038db41 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-20T05:33:06Z |
| publishDate | 2020-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-3248186172f54b829d379f64d038db412022-12-21T19:51:41ZengIEEEIEEE Access2169-35362020-01-01816945616946910.1109/ACCESS.2020.30239499195865Binary Imbalanced Data Classification Based on Modified D2GAN Oversampling and Classifier FusionJunhai Zhai0https://orcid.org/0000-0001-9962-7417Jiaxing Qi1Sufang Zhang2https://orcid.org/0000-0002-7585-6490Hebei Key Laboratory of Machine Learning and Computational Intelligence, College of Mathematics and Information Science, Hebei University, Baoding, ChinaHebei Key Laboratory of Machine Learning and Computational Intelligence, College of Mathematics and Information Science, Hebei University, Baoding, ChinaHebei Branch of China Meteorological Administration Training Center, China Meteorological Administration, Baoding, ChinaBinary imbalance problem refers to such a classification scenario where one class contains a large number of samples while another class contains only a few samples. When traditional classifiers face with imbalanced datasets, they usually bias towards majority class resulting in poor classification performance. Oversampling is an effective method to address this problem, yet how to conduct diversity oversampling is a challenge. In this article, we proposed a diversity oversampling method based on a modified D2GAN model, and on the basis of diversity oversampling, we also proposed a binary imbalanced data classification approach based on classifier fusion by fuzzy integral. Extensive experiments are conducted on 8 data sets to compare the proposed methods with 7 state-of-the-art methods on 5 aspects: MMD-score, Silhouette-score, F-measure, G-means, and AUC-area. The 7 methods include 4 SMOTE related approaches and 3 GAN related approaches. The experimental results demonstrate that the proposed methods are more effective and efficient than the compared approaches.https://ieeexplore.ieee.org/document/9195865/Binary class imbalancediversity oversamplinggenerative adversarial networkclassifier fusionfuzzy integral |
| spellingShingle | Junhai Zhai Jiaxing Qi Sufang Zhang Binary Imbalanced Data Classification Based on Modified D2GAN Oversampling and Classifier Fusion IEEE Access Binary class imbalance diversity oversampling generative adversarial network classifier fusion fuzzy integral |
| title | Binary Imbalanced Data Classification Based on Modified D2GAN Oversampling and Classifier Fusion |
| title_full | Binary Imbalanced Data Classification Based on Modified D2GAN Oversampling and Classifier Fusion |
| title_fullStr | Binary Imbalanced Data Classification Based on Modified D2GAN Oversampling and Classifier Fusion |
| title_full_unstemmed | Binary Imbalanced Data Classification Based on Modified D2GAN Oversampling and Classifier Fusion |
| title_short | Binary Imbalanced Data Classification Based on Modified D2GAN Oversampling and Classifier Fusion |
| title_sort | binary imbalanced data classification based on modified d2gan oversampling and classifier fusion |
| topic | Binary class imbalance diversity oversampling generative adversarial network classifier fusion fuzzy integral |
| url | https://ieeexplore.ieee.org/document/9195865/ |
| work_keys_str_mv | AT junhaizhai binaryimbalanceddataclassificationbasedonmodifiedd2ganoversamplingandclassifierfusion AT jiaxingqi binaryimbalanceddataclassificationbasedonmodifiedd2ganoversamplingandclassifierfusion AT sufangzhang binaryimbalanceddataclassificationbasedonmodifiedd2ganoversamplingandclassifierfusion |