Selective ensemble method for anomaly detection based on parallel learning
Abstract Anomaly detection is a highly important task in the field of data analysis. Traditional anomaly detection approaches often strongly depend on data size, structure and features, while introducing the idea of ensemble into anomaly detection can greatly improve the generalization ability. Ense...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2024-01-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-024-51849-3 |
| _version_ | 1797349946942291968 |
|---|---|
| author | Yansong Liu Li Zhu Lei Ding Zifeng Huang He Sui Shuang Wang Yuedong Song |
| author_facet | Yansong Liu Li Zhu Lei Ding Zifeng Huang He Sui Shuang Wang Yuedong Song |
| author_sort | Yansong Liu |
| collection | DOAJ |
| description | Abstract Anomaly detection is a highly important task in the field of data analysis. Traditional anomaly detection approaches often strongly depend on data size, structure and features, while introducing the idea of ensemble into anomaly detection can greatly improve the generalization ability. Ensemble-based anomaly detection methods still face some challenges, however, such as data imbalance, time and space demand and the selection of base detectors. To this end, we propose a selective ensemble method for anomaly detection based on parallel learning (SEAD-PL). First, a differentiated stratified sampling method is designed to alleviate the problem of data imbalance. Then, a distributed parallel training frame is built to address the problem of excessive time and space consumption for base detector training. Finally, a clustering-based ensemble selection strategy is introduced to balance the accuracy and diversity of base detectors. Experiments are performed on six datasets, which demonstrate that the proposed method has obvious advantages over four selected methods. |
| first_indexed | 2024-03-08T12:38:42Z |
| format | Article |
| id | doaj.art-b2d0d994acde47dcb339a968a6d60c4c |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-03-08T12:38:42Z |
| publishDate | 2024-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-b2d0d994acde47dcb339a968a6d60c4c2024-01-21T12:20:24ZengNature PortfolioScientific Reports2045-23222024-01-0114112110.1038/s41598-024-51849-3Selective ensemble method for anomaly detection based on parallel learningYansong Liu0Li Zhu1Lei Ding2Zifeng Huang3He Sui4Shuang Wang5Yuedong Song6School of Software Engineering, Xi’an Jiao Tong UniversitySchool of Software Engineering, Xi’an Jiao Tong UniversitySchool of Cyberspace Security, Guangzhou UniversitySchool of Electronics and Communication Engineering, Guangzhou UniversityCollege of Aeronautical Engineering, Civil Aviation University of ChinaInformation Security Evaluation Center of Civil Aviation, Civil Aviation University of ChinaShanghai Hua Xun Network Information System Co., LtdAbstract Anomaly detection is a highly important task in the field of data analysis. Traditional anomaly detection approaches often strongly depend on data size, structure and features, while introducing the idea of ensemble into anomaly detection can greatly improve the generalization ability. Ensemble-based anomaly detection methods still face some challenges, however, such as data imbalance, time and space demand and the selection of base detectors. To this end, we propose a selective ensemble method for anomaly detection based on parallel learning (SEAD-PL). First, a differentiated stratified sampling method is designed to alleviate the problem of data imbalance. Then, a distributed parallel training frame is built to address the problem of excessive time and space consumption for base detector training. Finally, a clustering-based ensemble selection strategy is introduced to balance the accuracy and diversity of base detectors. Experiments are performed on six datasets, which demonstrate that the proposed method has obvious advantages over four selected methods.https://doi.org/10.1038/s41598-024-51849-3 |
| spellingShingle | Yansong Liu Li Zhu Lei Ding Zifeng Huang He Sui Shuang Wang Yuedong Song Selective ensemble method for anomaly detection based on parallel learning Scientific Reports |
| title | Selective ensemble method for anomaly detection based on parallel learning |
| title_full | Selective ensemble method for anomaly detection based on parallel learning |
| title_fullStr | Selective ensemble method for anomaly detection based on parallel learning |
| title_full_unstemmed | Selective ensemble method for anomaly detection based on parallel learning |
| title_short | Selective ensemble method for anomaly detection based on parallel learning |
| title_sort | selective ensemble method for anomaly detection based on parallel learning |
| url | https://doi.org/10.1038/s41598-024-51849-3 |
| work_keys_str_mv | AT yansongliu selectiveensemblemethodforanomalydetectionbasedonparallellearning AT lizhu selectiveensemblemethodforanomalydetectionbasedonparallellearning AT leiding selectiveensemblemethodforanomalydetectionbasedonparallellearning AT zifenghuang selectiveensemblemethodforanomalydetectionbasedonparallellearning AT hesui selectiveensemblemethodforanomalydetectionbasedonparallellearning AT shuangwang selectiveensemblemethodforanomalydetectionbasedonparallellearning AT yuedongsong selectiveensemblemethodforanomalydetectionbasedonparallellearning |