Abstract:Technological advancements in cryptocurrency markets have increased accessibility for investors, but concurrently exposed them to the risks of market manipulations. Existing fraud detection mechanisms typically rely on machine learning methods that treat each financial asset (i.e., token) and its related transactions independently. However, market manipulation strategies are rarely isolated events, but are rather characterized by coordination, repetition, and frequent transfers among related assets. This suggests that relational structure constitutes an integral component of the signal and can be effectively represented through graphical means. In this paper, we propose three graph construction methods that rely on aggregated hourly market data. The proposed graphs are processed by a unified spatio-temporal Graph Neural Network (GNN) architecture that combines attention-based spatial aggregation with temporal Transformer encoding. We evaluate our methodology on a real-world dataset comprised of pump-and-dump schemes in cryptocurrency markets, spanning a period of over three years. Our comparative results showcase that our graph-based models achieve significant improvements over standard machine learning baselines in detecting anomalous events. Our work highlights that learned market connectivity provides substantial gains for detecting coordinated market manipulation schemes.
| Comments: | 9 pages, 3 figures, Accepted at the SDS2026: IEEE Swiss Conference on Data Science and AI |
| Subjects: | Machine Learning (cs.LG); Computational Engineering, Finance, and Science (cs.CE) |
| ACM classes: | I.2.6; K.4.4; G.2.2 |
| Cite as: | arXiv:2604.24590 [cs.LG] |
| (or arXiv:2604.24590v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2604.24590 arXiv-issued DOI via DataCite (pending registration) |
From: Dimosthenis Pasadakis [view email]
[v1]
Mon, 27 Apr 2026 15:12:11 UTC (10,970 KB)
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。