arXiv 2503.179032025-12-26cs.LGcs.AI

GLADMamba: 一种基于选择性状态空间模型的无监督图级异常检测方法

GLADMamba: Unsupervised Graph-Level Anomaly Detection Powered by Selective State Space Model

School of Artificial Intelligence, Jilin University, Changchun, China（吉林大学人工智能学院）；Hong Kong University of Science and Technology (Guangzhou), Guangzhou, China（香港科技大学（广州））；College of Computer Science and Technology, Jilin University, Changchun, China（吉林大学计算机科学与技术学院）；Engineering Research Center of Knowledge-Driven Human-Machine Intelligence, Ministry of Education, China（知识驱动的人机智能工程研究中心，教育部）

Yali Fu, Jindong Li, Qi Wang, Qianli Xing

备注：: Published at ECML PKDD 2025 (Research Track)
DOI：: 10.1007/978-3-032-05962-8_3

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

GLADMamba通过引入选择性状态空间模型和谱信息，实现了无监督图级异常检测的高效检测与优化。

摘要：

无监督图级异常检测（UGLAD）是跨多个领域（如社交网络分析、抗癌药物发现和有毒分子识别）的关键且具有挑战性的任务。然而，现有方法往往难以高效捕捉长距离依赖关系，并且忽视了谱信息。最近，选择性状态空间模型，特别是Mamba，已在捕捉长距离依赖关系方面表现出显著优势，具有线性复杂度和选择机制。受其在不同领域中的成功启发，我们提出GLADMamba，一种将选择性状态空间模型引入UGLAD领域的新型框架。我们设计了一个视图融合Mamba（VFM）模块，采用Mamba-Transformer式架构，以选择性状态机制高效融合不同图视图的信息。我们还设计了一个谱引导Mamba（SGM）模块，采用Mamba-Transformer式架构，利用Rayleigh商来引导嵌入细化过程，考虑谱信息以实现UGLAD。GLADMamba可以动态聚焦于与异常相关的信息，同时丢弃无关信息以进行异常检测。据我们所知，这是首次将Mamba和显式谱信息引入UGLAD的研究。在12个真实世界数据集上的广泛实验表明，GLADMamba优于现有最先进方法，在UGLAD中实现了更优的性能。代码可在https://github.com/Yali-Fu/GLADMamba获取。

英文摘要：

Unsupervised graph-level anomaly detection (UGLAD) is a critical and challenging task across various domains, such as social network analysis, anti-cancer drug discovery, and toxic molecule identification. However, existing methods often struggle to capture long-range dependencies efficiently and neglect the spectral information. Recently, selective state space models, particularly Mamba, have demonstrated remarkable advantages in capturing long-range dependencies with linear complexity and a selection mechanism. Motivated by their success across various domains, we propose GLADMamba, a novel framework that adapts the selective state space model into UGLAD field. We design a View-Fused Mamba (VFM) module with a Mamba-Transformer-style architecture to efficiently fuse information from different graph views with a selective state mechanism. We also design a Spectrum-Guided Mamba (SGM) module with a Mamba-Transformer-style architecture to leverage the Rayleigh quotient to guide the embedding refinement process, considering the spectral information for UGLAD. GLADMamba can dynamically focus on anomaly-related information while discarding irrelevant information for anomaly detection. To the best of our knowledge, this is the first work to introduce Mamba and explicit spectral information to UGLAD. Extensive experiments on 12 real-world datasets demonstrate that GLADMamba outperforms existing state-of-the-art methods, achieving superior performance in UGLAD. The code is available at https://github.com/Yali-Fu/GLADMamba.

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。