后验状态与潜在欺骗：隐马尔可夫模型中的变分贝叶斯推断在金融交易序列欺诈检测中的应用
Posterior Regimes and Latent Deception: Variational Bayesian Inference in Hidden Markov Models for Sequential Fraud Detection in Financial Transactions
浏览论文内容
中文总结 AI 辅助
本文提出三层隐马尔可夫模型（最大似然、变分贝叶斯、神经变分）用于金融交易序列欺诈检测，通过潜在状态识别欺诈，并指出后验概率常被误用，提出校准的分诊层而非直接替代排序模型。
中文摘要 AI 辅助
我们提出了隐马尔可夫模型的三层递进结构：最大似然（Baum-Welch）、变分贝叶斯（VBEM）以及神经变分扩展（Neural VBEM），将每位客户的交易历史建模为通过少量潜在行为状态的轨迹，其中一种状态被经验性地识别为与欺诈相关。Neural VBEM HMM 用学习到的编码器替换了固定的高斯-多项发射族，将 741 维的交易表示压缩到 64 维的潜在空间，在该空间中 VBEM HMM 的后验进行操作；该空间的 UMAP 投影揭示，发现的状态并非离散簇，而是单一连续行为流形上的有序片段，已确认的欺诈集中在流形的极端。我们表明，模型的自然输出，即状态成员的后验概率，常被误认为是欺诈概率，并量化了由此产生的校准误差（状态成员解释误差，MRIE）；修正后的后验预测分数缩小了大部分差距。我们进一步区分了批量（平滑）推断（使用部署时不可用的前瞻信息）和滤波（仅前向）推断，并报告两者。在 IEEE-CIS 交易数据上，神经层在其识别状态中实现了 14.4 倍的欺诈富集；虽然其 AUPRC 落后于判别式 XGBoost 基线，但我们表明这一差距是结构性的而非偶然的，并认为该模型最适合作为校准的分诊和可解释性层，而非直接替代的排序模型。
英文摘要
We present a three-tier progression of Hidden Markov Models: maximum-likelihood (Baum-Welch), variational Bayesian (VBEM), and a neural variational extension (Neural VBEM), that model each customer's transaction history as a trajectory through a small number of latent behavioural regimes, one of which is empirically identified as fraud-associated. The Neural VBEM HMM replaces the fixed Gaussian-multinomial emission family with a learned encoder, compressing a 741-dimensional transaction representation into a 64-dimensional latent space in which the VBEM HMM's posterior operates; a UMAP projection of this space reveals that the discovered regimes are not discrete clusters but ordered segments of a single continuous behavioural manifold, with confirmed fraud concentrated at its extreme. We show that the model's natural output, that is, the posterior probability of regime membership, is routinely mistaken for a fraud probability, and quantify the resulting miscalibration (the regime-membership interpretation error, MRIE); a corrected posterior-predictive score, closes most of this gap. We further distinguish batch (smoothed) inference, which uses look-ahead unavailable at deployment time, from filtered (forward-only) inference, and report both. On IEEE-CIS transaction data, the neural tier achieves a 14.4$\times$ fraud enrichment in its identified regime; while its AUPRC trails a discriminative XGBoost baseline, we show this gap is structural and not incidental, and argue the model is best positioned as a calibrated triage and interpretability layer rather than a drop-in ranking replacement.
发表机构
- University of Stirling（斯特灵大学）
- University of Nottingham（诺丁汉大学）
机构由 AI 辅助整理，请以论文原文为准。
