Feature Selection with Conditional Mutual Information Considering Feature Interaction
Feature interaction is a newly proposed feature relevance relationship, but the unintentional removal of interactive features can result in poor classification performance for this relationship. However, traditional feature selection algorithms mainly focus on detecting relevant and redundant featur...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2019-07-01
|
| Series: | Symmetry |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-8994/11/7/858 |
| _version_ | 1811187883126554624 |
|---|---|
| author | Jun Liang Liang Hou Zhenhua Luan Weiping Huang |
| author_facet | Jun Liang Liang Hou Zhenhua Luan Weiping Huang |
| author_sort | Jun Liang |
| collection | DOAJ |
| description | Feature interaction is a newly proposed feature relevance relationship, but the unintentional removal of interactive features can result in poor classification performance for this relationship. However, traditional feature selection algorithms mainly focus on detecting relevant and redundant features while interactive features are usually ignored. To deal with this problem, feature relevance, feature redundancy and feature interaction are redefined based on information theory. Then a new feature selection algorithm named CMIFSI (Conditional Mutual Information based Feature Selection considering Interaction) is proposed in this paper, which makes use of conditional mutual information to estimate feature redundancy and interaction, respectively. To verify the effectiveness of our algorithm, empirical experiments are conducted to compare it with other several representative feature selection algorithms. The results on both synthetic and benchmark datasets indicate that our algorithm achieves better results than other methods in most cases. Further, it highlights the necessity of dealing with feature interaction. |
| first_indexed | 2024-04-11T14:11:07Z |
| format | Article |
| id | doaj.art-47094849007f4ce497f1a95d17d5a1b3 |
| institution | Directory Open Access Journal |
| issn | 2073-8994 |
| language | English |
| last_indexed | 2024-04-11T14:11:07Z |
| publishDate | 2019-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Symmetry |
| spelling | doaj.art-47094849007f4ce497f1a95d17d5a1b32022-12-22T04:19:42ZengMDPI AGSymmetry2073-89942019-07-0111785810.3390/sym11070858sym11070858Feature Selection with Conditional Mutual Information Considering Feature InteractionJun Liang0Liang Hou1Zhenhua Luan2Weiping Huang3State Key Lab of Nuclear Power Safety Monitoring Technology and Equipment, Shenzhen 518124, ChinaState Key Lab of Industrial Control Technology, College of Control Science and Engineering, Zhejiang University, Hangzhou 310027, ChinaState Key Lab of Nuclear Power Safety Monitoring Technology and Equipment, Shenzhen 518124, ChinaState Key Lab of Industrial Control Technology, College of Control Science and Engineering, Zhejiang University, Hangzhou 310027, ChinaFeature interaction is a newly proposed feature relevance relationship, but the unintentional removal of interactive features can result in poor classification performance for this relationship. However, traditional feature selection algorithms mainly focus on detecting relevant and redundant features while interactive features are usually ignored. To deal with this problem, feature relevance, feature redundancy and feature interaction are redefined based on information theory. Then a new feature selection algorithm named CMIFSI (Conditional Mutual Information based Feature Selection considering Interaction) is proposed in this paper, which makes use of conditional mutual information to estimate feature redundancy and interaction, respectively. To verify the effectiveness of our algorithm, empirical experiments are conducted to compare it with other several representative feature selection algorithms. The results on both synthetic and benchmark datasets indicate that our algorithm achieves better results than other methods in most cases. Further, it highlights the necessity of dealing with feature interaction.https://www.mdpi.com/2073-8994/11/7/858feature selectionconditional mutual informationfeature interactionclassificationcomputer engineering |
| spellingShingle | Jun Liang Liang Hou Zhenhua Luan Weiping Huang Feature Selection with Conditional Mutual Information Considering Feature Interaction Symmetry feature selection conditional mutual information feature interaction classification computer engineering |
| title | Feature Selection with Conditional Mutual Information Considering Feature Interaction |
| title_full | Feature Selection with Conditional Mutual Information Considering Feature Interaction |
| title_fullStr | Feature Selection with Conditional Mutual Information Considering Feature Interaction |
| title_full_unstemmed | Feature Selection with Conditional Mutual Information Considering Feature Interaction |
| title_short | Feature Selection with Conditional Mutual Information Considering Feature Interaction |
| title_sort | feature selection with conditional mutual information considering feature interaction |
| topic | feature selection conditional mutual information feature interaction classification computer engineering |
| url | https://www.mdpi.com/2073-8994/11/7/858 |
| work_keys_str_mv | AT junliang featureselectionwithconditionalmutualinformationconsideringfeatureinteraction AT lianghou featureselectionwithconditionalmutualinformationconsideringfeatureinteraction AT zhenhualuan featureselectionwithconditionalmutualinformationconsideringfeatureinteraction AT weipinghuang featureselectionwithconditionalmutualinformationconsideringfeatureinteraction |