Fairness with Adaptive Weight in Network Attack Detection

Xuezhi Wen,Nan Wang,Yuanlin Sun,Fan Xu,Dalin Zhang,Xibin Zhao
DOI: https://doi.org/10.1109/icpads60453.2023.00190
2023-01-01
Abstract:Network attacks aim to exploit vulnerabilities inherent in network protocols, which is widely used in many real-world applications. In the process of network anomaly detection, most methods train the model by minimizing the average empirical risk of all samples. However, due to the uneven distribution of samples from different protocols, detection models tend to be biased against minority protocols groups. To address this issue, we propose an adaptive weight assignment method for network attack detection, which emphasizes more on error-prone samples in prediction and enhances adequate representation of minority groups for fairness. We conduct experiments on two widely used datasets, KDD and NSLKDD. According to the results, our method achieves better performance than state-of-the-art methods for classification and regression tasks, and is robust to label noise in the test dataset.
What problem does this paper attempt to address?