原文传递
Competitively Learned Attention Mechanism Prototypes for Network Intrusion Detection.
| 题名： | Competitively Learned Attention Mechanism Prototypes for Network Intrusion Detection. |
| 作者： | Heckel, K. M. |
| 摘要： | Maintaining secure computer networks and information systems is critical to the administrative functions and operations carried out by the Department ofDefense (DOD). Ensuring the security and functionality of these networks is vital to US National Security, however, previous work on applying neuralnetworks to intrusion detection has focused on recurrent and convolutional neural networks but has yet to explore attention-mechanism-based architectures.Inspired by the Transformer in Vaswani et al. (2017), These attention-based models rely predominantly on layers that produce rich contextualizedrepresentations through learning pairwise interactions within sequences, enabling significant advances in computer vision and natural language processingover the last five years.This research investigated the performance of attention-based neural network architectures compared to traditional models on the University of NewBrunswick’s CSE-CIC-IDS2018 dataset. Reporting precision, recall, and the Area Under the Receiver Operating Characteristic Curve (AUROC), resultsshow that models leveraging attention mechanisms performed demonstrably better than a tuned feed-forward network on the infiltration attack class.Additionally, this work explores a novel efficient attention mechanism for deep neural networks by learning a compressed representation of the datathrough competitively-learned memory prototypes, showing competitive performance against an alternative efficient attention architecture that utilizesgradient descent. |
| 总页数： | 42 pages |
相关文献
- IMPROVING SURVIVABILITY OF INTELLIGENT TRANSPORTATION SYSTEMS: APPLICATION-LEVEL INTRUSION DETECTION CASE STUDY.
- Cyber Data Anomaly Detection Using Autoencoder Neural Networks.
- Use of Packet Header Anomaly Detection in Lossy Network Traffic Compression for Network Intrusion Detection Applications.
- Methods to Address Extreme Class Imbalance in Machine Learning Based Network Intrusion Detection Systems.
- Critical Energy Infrastructure Cyber Defense-in-Depth.
- Evaluation of Intrusion Detection Technologies for High Speed Rail Grade Crossings.
检索历史
应用推荐
