arXiv cs.CY· Seongchan Park, Hyunho Jeon, Byungsoo Kang, Gaeun Kim, Minseob So, Seongyong Kim·· 11 小时前AI 评分31
基于储备证明的以太坊中心化交易所地址链上行为特征分类框架
On-Chain Behavioral Signatures for Classifying Centralized Exchange Addresses on Ethereum Using Proof-of-Reserves
AI 导读
研究利用储备证明（PoR）披露信息构建分类框架，从15家中心化交易所（CEX）和五类非CEX来源的3,033,857个以太坊地址中提取58个链上行为特征，在留一交易所交叉验证下LightGBM的macro PR-AUC达0.9594，优于逻辑回归的0.7697。SHAP分析显示gas行为、ERC-20 token活动和内部交易是关键预测信号，该框架支持可解释的CEX地址筛查并已公开特征数据集。
正文 · 原文
Abstract:Centralized exchanges (CEXs) are key gateways between traditional finance and digital asset markets, yet identifying their blockchain addresses remains challenging. Existing approaches often rely on opaque attribution labels or treat exchanges as secondary classification targets, leaving generalization to unseen exchanges insufficiently examined. This study develops a classification framework using Proof-of-Reserves (PoR) disclosures, with 3,033,857 Ethereum addresses from 15 CEXs and five Non-CEX categories. We derive 58 on-chain behavioral features and evaluate LightGBM against logistic regression (LR) under a leave-one-exchange-out protocol. Across five lookback windows (30-730 days), LightGBM consistently outperformed LR, with macro PR-AUC increasing from 0.6611 to 0.9594 using the selected 17 features, compared with 0.7697 for LR at 730 days. Robustness checks showed comparable performance when feature selection and hyperparameter optimization were repeated across splits and held-out exchanges, while exchange-balanced reweighting improved macro PR-AUC. SHAP analysis and feature-group ablation identified gas behavior, ERC-20 token activity, and internal transactions as key predictive signals in the constructed dataset. These findings support interpretable CEX address screening for blockchain monitoring, while the released feature dataset enables reproducible follow-up research.
| Comments: | Accepted at the 7th ACM International Conference on AI in Finance (ICAIF '26) |
| Subjects: | Computers and Society (cs.CY); Cryptography and Security (cs.CR); Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.41888 [cs.CY] |
| (or arXiv:2610.41888v1 [cs.CY] for this version) | |
| https://doi.org/10.48550/arXiv.2610.41888 arXiv-issued DOI via DataCite (pending registration) |
|
| Related DOI: | https://doi.org/10.1145/3858063.3858815
DOI(s) linking to related resources |
Submission history
From: Seongchan Park [view email]
[v1]
Fri, 30 Oct 2026 15:32:04 UTC (487 KB)
来源：arXiv cs.CY · arxiv.org