DRnet: Dynamic Retraining for Malicious Traffic Small-Sample Incremental Learning
Deep learning has achieved good classification results in the field of traffic classification in recent years due to its good feature representation ability. However, the existing traffic classification technology cannot meet the requirements for the incremental learning of tasks in online scenarios...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-06-01
|
| Series: | Electronics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2079-9292/12/12/2668 |
| _version_ | 1797595142536822784 |
|---|---|
| author | Ruonan Wang Jinlong Fei Rongkai Zhang Maohua Guo Zan Qi Xue Li |
| author_facet | Ruonan Wang Jinlong Fei Rongkai Zhang Maohua Guo Zan Qi Xue Li |
| author_sort | Ruonan Wang |
| collection | DOAJ |
| description | Deep learning has achieved good classification results in the field of traffic classification in recent years due to its good feature representation ability. However, the existing traffic classification technology cannot meet the requirements for the incremental learning of tasks in online scenarios. In addition, due to the high concealment and fast update speed of malicious traffic, the number of labeled samples that can be captured is scarce, and small samples cannot drive neural network training, resulting in poor performance of the classification model. Therefore, this paper proposes an incremental learning method for small-sample malicious traffic classification. The method uses the pruning strategy to find the redundant network structure and dynamically allocates redundant neurons for training based on the proposed measurement method according to the difficulty of the new class. This enables the network to perform incremental learning without excessively consuming storage and computing resources, and reasonable allocation improves the classification accuracy of new classes. At the same time, through the knowledge transfer method, the model can reduce the catastrophic forgetting of the old class, relieve the pressure of training large parameters with small-sample data, and improve the model classification performance. Experiments involving multiple datasets and settings show that our method is superior to the established baseline in terms of classification accuracy, consuming 50% less memory. |
| first_indexed | 2024-03-11T02:33:14Z |
| format | Article |
| id | doaj.art-f739256efb30407e98425c1bd24ddf08 |
| institution | Directory Open Access Journal |
| issn | 2079-9292 |
| language | English |
| last_indexed | 2024-03-11T02:33:14Z |
| publishDate | 2023-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Electronics |
| spelling | doaj.art-f739256efb30407e98425c1bd24ddf082023-11-18T10:08:57ZengMDPI AGElectronics2079-92922023-06-011212266810.3390/electronics12122668DRnet: Dynamic Retraining for Malicious Traffic Small-Sample Incremental LearningRuonan Wang0Jinlong Fei1Rongkai Zhang2Maohua Guo3Zan Qi4Xue Li5State Key Laboratory of Mathematical Engineering and Advanced Computing, PLA Strategic Support Force Information Engineering University, Zhengzhou 450001, ChinaState Key Laboratory of Mathematical Engineering and Advanced Computing, PLA Strategic Support Force Information Engineering University, Zhengzhou 450001, ChinaState Key Laboratory of Mathematical Engineering and Advanced Computing, PLA Strategic Support Force Information Engineering University, Zhengzhou 450001, ChinaState Key Laboratory of Mathematical Engineering and Advanced Computing, PLA Strategic Support Force Information Engineering University, Zhengzhou 450001, ChinaState Key Laboratory of Mathematical Engineering and Advanced Computing, PLA Strategic Support Force Information Engineering University, Zhengzhou 450001, ChinaState Key Laboratory of Mathematical Engineering and Advanced Computing, PLA Strategic Support Force Information Engineering University, Zhengzhou 450001, ChinaDeep learning has achieved good classification results in the field of traffic classification in recent years due to its good feature representation ability. However, the existing traffic classification technology cannot meet the requirements for the incremental learning of tasks in online scenarios. In addition, due to the high concealment and fast update speed of malicious traffic, the number of labeled samples that can be captured is scarce, and small samples cannot drive neural network training, resulting in poor performance of the classification model. Therefore, this paper proposes an incremental learning method for small-sample malicious traffic classification. The method uses the pruning strategy to find the redundant network structure and dynamically allocates redundant neurons for training based on the proposed measurement method according to the difficulty of the new class. This enables the network to perform incremental learning without excessively consuming storage and computing resources, and reasonable allocation improves the classification accuracy of new classes. At the same time, through the knowledge transfer method, the model can reduce the catastrophic forgetting of the old class, relieve the pressure of training large parameters with small-sample data, and improve the model classification performance. Experiments involving multiple datasets and settings show that our method is superior to the established baseline in terms of classification accuracy, consuming 50% less memory.https://www.mdpi.com/2079-9292/12/12/2668malicious traffic classificationsmall samplesincremental learningdynamic retraining |
| spellingShingle | Ruonan Wang Jinlong Fei Rongkai Zhang Maohua Guo Zan Qi Xue Li DRnet: Dynamic Retraining for Malicious Traffic Small-Sample Incremental Learning Electronics malicious traffic classification small samples incremental learning dynamic retraining |
| title | DRnet: Dynamic Retraining for Malicious Traffic Small-Sample Incremental Learning |
| title_full | DRnet: Dynamic Retraining for Malicious Traffic Small-Sample Incremental Learning |
| title_fullStr | DRnet: Dynamic Retraining for Malicious Traffic Small-Sample Incremental Learning |
| title_full_unstemmed | DRnet: Dynamic Retraining for Malicious Traffic Small-Sample Incremental Learning |
| title_short | DRnet: Dynamic Retraining for Malicious Traffic Small-Sample Incremental Learning |
| title_sort | drnet dynamic retraining for malicious traffic small sample incremental learning |
| topic | malicious traffic classification small samples incremental learning dynamic retraining |
| url | https://www.mdpi.com/2079-9292/12/12/2668 |
| work_keys_str_mv | AT ruonanwang drnetdynamicretrainingformalicioustrafficsmallsampleincrementallearning AT jinlongfei drnetdynamicretrainingformalicioustrafficsmallsampleincrementallearning AT rongkaizhang drnetdynamicretrainingformalicioustrafficsmallsampleincrementallearning AT maohuaguo drnetdynamicretrainingformalicioustrafficsmallsampleincrementallearning AT zanqi drnetdynamicretrainingformalicioustrafficsmallsampleincrementallearning AT xueli drnetdynamicretrainingformalicioustrafficsmallsampleincrementallearning |