Uncertainty-aware Human Mobility Modeling and Anomaly Detection

TL;DR

This paper introduces USTAD, an unsupervised, uncertainty-aware model for human mobility anomaly detection using GPS data, which effectively accounts for behavioral stochasticity and data sparsity.

Contribution

The work presents a novel sequence modeling approach with uncertainty estimation for anomaly detection in human mobility without labeled data.

Findings

USTAD improves anomaly detection AUCROC by 3-15% over baselines.

Incorporates aleatoric and epistemic uncertainties for robustness.

Effective on industry-scale GPS data.

Abstract

Given the temporal GPS coordinates from a large set of human agents, how can we model their mobility behavior toward effective anomaly (e.g. bad-actor or malicious behavior) detection without any labeled data? Human mobility and trajectory modeling have been extensively studied, showcasing varying abilities to manage complex inputs and balance performance-efficiency trade-offs. In this work, we formulate anomaly detection in complex human behavior by modeling raw GPS data as a sequence of stay-point events, each characterized by spatio-temporal features, along with trips (i.e. commute) between the stay-points. Our problem formulation allows us to leverage modern sequence models for unsupervised training and anomaly detection. Notably, we equip our proposed model USTAD (for Uncertainty-aware Spatio-Temporal Anomaly Detection) with aleatoric (i.e. data) uncertainty estimation to account for inherent stochasticity in certain individuals' behavior, as well as epistemic (i.e. model) uncertainty to handle data sparsity under a large variety of human behaviors. Together, aleatoric and epistemic uncertainties unlock a robust loss function as well as uncertainty-aware decision-making in anomaly scoring. Extensive experiments shows that USTAD improves anomaly detection AUCROC by 3\%-15\% over baselines in industry-scale data.