On Multi-entity, Multivariate Quickest Change Point Detection

2025-09-22

AI Summary

This paper addresses the problem of online system-level change-point detection in complex dynamic environments involving multivariate time series from multiple entities. The proposed method introduces an unsupervised, feature-engineering-free real-time detection framework: it defines an individual anomaly metric (IDfN) based on autoencoder reconstruction error and a system-wide anomaly score (SWAS) that integrates mean, variance, and kernel density estimation; robust change-point localization is achieved via CUSUM. Key contributions include: (i) the first benchmark dataset for group-behavior abrupt-change detection over multi-entity time series; (ii) a scalable, privacy-preserving detection paradigm; and (iii) extensive evaluation on synthetic and Unity-based crowd-simulation data, demonstrating superior accuracy and robustness compared to conventional monitoring approaches.

Technology Category

• Data Mining & Knowledge Management: Anomaly/Outlier Detection
• Machine Learning: Time-Series/Data Streams
• Humans and AI: Crowd Sourcing and Human Computation

Application Category

• Security and Privacy: Large-scale security measurements
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Economics, Online Markets and Human Computation: Fairness, privacy, and diversity in economic environments

Abstract

We propose a framework for online Change Point Detection (CPD) from multi-entity, multivariate time series data, motivated by applications in crowd monitoring where traditional sensing methods (e.g., video surveillance) may be infeasible. Our approach addresses the challenge of detecting system-wide behavioral shifts in complex, dynamic environments where the number and behavior of individual entities may be uncertain or evolve. We introduce the concept of Individual Deviation from Normality (IDfN), computed via a reconstruction-error-based autoencoder trained on normal behavior. We aggregate these individual deviations using mean, variance, and Kernel Density Estimates (KDE) to yield a System-Wide Anomaly Score (SWAS). To detect persistent or abrupt changes, we apply statistical deviation metrics and the Cumulative Sum (CUSUM) technique to these scores. Our unsupervised approach eliminates the need for labeled data or feature extraction, enabling real-time operation on streaming input. Evaluations on both synthetic datasets and crowd simulations, explicitly designed for anomaly detection in group behaviors, demonstrate that our method accurately detects significant system-level changes, offering a scalable and privacy-preserving solution for monitoring complex multi-agent systems. In addition to this methodological contribution, we introduce new, challenging multi-entity multivariate time series datasets generated from crowd simulations in Unity and coupled nonlinear oscillators. To the best of our knowledge, there is currently no publicly available dataset of this type designed explicitly to evaluate CPD in complex collective and interactive systems, highlighting an essential gap that our work addresses.

Problem

• Detecting system-wide behavioral shifts in multi-entity multivariate time series
• Monitoring complex dynamic environments with uncertain evolving entity behavior
• Providing unsupervised real-time change detection without labeled data requirements

Innovation

• Autoencoder-based Individual Deviation from Normality (IDfN)
• System-Wide Anomaly Score (SWAS) via aggregated statistics
• CUSUM technique applied to scores for change detection