Adaptive KDE for Real-Time Thresholding: Prioritized Queues for Financial Crime Investigation
This addresses the need for real-time, label-free thresholding in financial crime investigation, though it appears incremental as it builds on existing density estimation techniques.
The paper tackles the problem of converting a stream of risk scores into review queues under intake constraints by using an online adaptive kernel density method to set thresholds, achieving competitive capacity adherence and reduced threshold jitter on synthetic, drifting, multimodal streams.
We study the problem of converting a stream of risk scores into one or more review queues under explicit intake constraints[cite: 6]. Instead of top-$K$ or manually tuned cutoffs, we fit an online adaptive kernel density to the score stream, transform the density into a tail-mass curve to meet capacity, and ``snap'' the resulting cut to a persistent density valley detected across bandwidths[cite: 7]. The procedure is label-free, supports multi-queue routing, and operates in real time with sliding windows or exponential forgetting[cite: 8]. On synthetic, drifting, multimodal streams, the method achieves competitive capacity adherence while reducing threshold jitter[cite: 9]. Updates cost $O(G)$ per event with constant memory per activity