arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2412.16447cs.LGcs.AI
收藏

动态图中的一种可泛化异常检测方法

A Generalizable Anomaly Detection Method in Dynamic Graphs

Nanyang Technological University (NTU)(南洋理工大学);Joint NTU-UBC Research Centre of Excellence in Active Living for the Elderly (LILY)(南洋理工大学-不列颠哥伦比亚大学活跃老年生活卓越联合研究中心(LILY));Joint NTU-WeBank Research Centre on Fintech(南洋理工大学-微众银行金融科技联合研究中心)

Xiao Yang, Xuejiao Zhao, Zhiqi Shen

AI总结:

针对动态图异常检测方法缺乏泛化能力的问题,提出GeneralDyG方法,通过采样时间自我图并顺序提取结构与时间特征,有效应对数据多样性、动态特征捕获与计算成本挑战,在四个真实数据集上显著优于现有方法。

摘要:

异常检测旨在识别数据中偏离正常模式的异常。由于动态图具有不断演化的结构和复杂的关系,在社交网络和网络安全等应用中极为常见,因此该任务在动态图中尤为重要。尽管近期基于深度学习的动态图异常检测方法取得了较好效果,但往往缺乏泛化能力。本研究提出GeneralDyG方法,通过采样时间自我图并顺序提取结构和时间特征,以解决实现泛化能力的三大关键挑战:数据多样性、动态特征捕获和计算成本。大量实验表明,所提GeneralDyG在四个真实世界数据集上显著优于现有最先进方法。

英文摘要:

Anomaly detection aims to identify deviations from normal patterns within data. This task is particularly crucial in dynamic graphs, which are common in applications like social networks and cybersecurity, due to their evolving structures and complex relationships. Although recent deep learning-based methods have shown promising results in anomaly detection on dynamic graphs, they often lack of generalizability. In this study, we propose GeneralDyG, a method that samples temporal ego-graphs and sequentially extracts structural and temporal features to address the three key challenges in achieving generalizability: Data Diversity, Dynamic Feature Capture, and Computational Cost. Extensive experimental results demonstrate that our proposed GeneralDyG significantly outperforms state-of-the-art methods on four real-world datasets.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑