Research and Application of Anormaly Detection Based on Improved DM-SVDD Algorithm
Aiming at the problem that traditional anomaly detection model has poor recognition effect on a small number of abnormal samples under the condition of data imbalance, in this paper we proposed a support vector data description algorithm combined with improved diffusion maps (DM-SVDD), constructed a...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Editorial Office of Journal of Taiyuan University of Technology
2021-09-01
|
| Series: | Taiyuan Ligong Daxue xuebao |
| Subjects: | |
| Online Access: | https://tyutjournal.tyut.edu.cn/englishpaper/show-325.html |
| _version_ | 1797216739759489024 |
|---|---|
| author | Jie WANG Xueying ZHANG Fenglian LI Haiwen DU Lijun YU Xiu MA |
| author_facet | Jie WANG Xueying ZHANG Fenglian LI Haiwen DU Lijun YU Xiu MA |
| author_sort | Jie WANG |
| collection | DOAJ |
| description | Aiming at the problem that traditional anomaly detection model has poor recognition effect on a small number of abnormal samples under the condition of data imbalance, in this paper we proposed a support vector data description algorithm combined with improved diffusion maps (DM-SVDD), constructed a new model and applied it to industry abnormal detection. The diffusion mapping algorithm was improved by introducing Euclidean distance and Mahalanobis distance to construct a new neighbor graph. Combined with support vector data description algorithm for modeling, the new model improved the recognition performance of normal samples, and the detection performance of abnormal samples was better than that from traditional models. Experimental data were selected of polysilicon ingot data sets. The results show that for an unbalanced data set formed by fewer abnormal samples, compared with traditional anomaly detection model, the model proposed in this paper can increase G-Mean optimally by 15.73% and F-Score optimally by 19.37%, which meet the requirements of industrial anomaly detection. The model can be used to guide the actual production process and reduce production costs. |
| first_indexed | 2024-04-24T11:50:45Z |
| format | Article |
| id | doaj.art-f27b44a50ee74e139a53a41a48c1d09b |
| institution | Directory Open Access Journal |
| issn | 1007-9432 |
| language | English |
| last_indexed | 2024-04-24T11:50:45Z |
| publishDate | 2021-09-01 |
| publisher | Editorial Office of Journal of Taiyuan University of Technology |
| record_format | Article |
| series | Taiyuan Ligong Daxue xuebao |
| spelling | doaj.art-f27b44a50ee74e139a53a41a48c1d09b2024-04-09T08:04:06ZengEditorial Office of Journal of Taiyuan University of TechnologyTaiyuan Ligong Daxue xuebao1007-94322021-09-0152576476810.16355/j.cnki.issn1007-9432tyut.2021.05.0101007-9432(2021)05-0764-05Research and Application of Anormaly Detection Based on Improved DM-SVDD AlgorithmJie WANG0Xueying ZHANG1Fenglian LI2Haiwen DU3Lijun YU4Xiu MA5College of Information and Computer, Taiyuan University of Technology, Taiyuan 030024, ChinaCollege of Information and Computer, Taiyuan University of Technology, Taiyuan 030024, ChinaCollege of Information and Computer, Taiyuan University of Technology, Taiyuan 030024, ChinaCETC New Energy Technology Co., Ltd, Taiyuan 030024, ChinaCETC New Energy Technology Co., Ltd, Taiyuan 030024, ChinaCETC New Energy Technology Co., Ltd, Taiyuan 030024, ChinaAiming at the problem that traditional anomaly detection model has poor recognition effect on a small number of abnormal samples under the condition of data imbalance, in this paper we proposed a support vector data description algorithm combined with improved diffusion maps (DM-SVDD), constructed a new model and applied it to industry abnormal detection. The diffusion mapping algorithm was improved by introducing Euclidean distance and Mahalanobis distance to construct a new neighbor graph. Combined with support vector data description algorithm for modeling, the new model improved the recognition performance of normal samples, and the detection performance of abnormal samples was better than that from traditional models. Experimental data were selected of polysilicon ingot data sets. The results show that for an unbalanced data set formed by fewer abnormal samples, compared with traditional anomaly detection model, the model proposed in this paper can increase G-Mean optimally by 15.73% and F-Score optimally by 19.37%, which meet the requirements of industrial anomaly detection. The model can be used to guide the actual production process and reduce production costs.https://tyutjournal.tyut.edu.cn/englishpaper/show-325.htmlsupport vector data descriptiondiffusion mapsanomaly detectionunbalanced datapolysilicon ingots data |
| spellingShingle | Jie WANG Xueying ZHANG Fenglian LI Haiwen DU Lijun YU Xiu MA Research and Application of Anormaly Detection Based on Improved DM-SVDD Algorithm Taiyuan Ligong Daxue xuebao support vector data description diffusion maps anomaly detection unbalanced data polysilicon ingots data |
| title | Research and Application of Anormaly Detection Based on Improved DM-SVDD Algorithm |
| title_full | Research and Application of Anormaly Detection Based on Improved DM-SVDD Algorithm |
| title_fullStr | Research and Application of Anormaly Detection Based on Improved DM-SVDD Algorithm |
| title_full_unstemmed | Research and Application of Anormaly Detection Based on Improved DM-SVDD Algorithm |
| title_short | Research and Application of Anormaly Detection Based on Improved DM-SVDD Algorithm |
| title_sort | research and application of anormaly detection based on improved dm svdd algorithm |
| topic | support vector data description diffusion maps anomaly detection unbalanced data polysilicon ingots data |
| url | https://tyutjournal.tyut.edu.cn/englishpaper/show-325.html |
| work_keys_str_mv | AT jiewang researchandapplicationofanormalydetectionbasedonimproveddmsvddalgorithm AT xueyingzhang researchandapplicationofanormalydetectionbasedonimproveddmsvddalgorithm AT fenglianli researchandapplicationofanormalydetectionbasedonimproveddmsvddalgorithm AT haiwendu researchandapplicationofanormalydetectionbasedonimproveddmsvddalgorithm AT lijunyu researchandapplicationofanormalydetectionbasedonimproveddmsvddalgorithm AT xiuma researchandapplicationofanormalydetectionbasedonimproveddmsvddalgorithm |