MIDAS: Microcluster-Based Detector of Anomalies in Edge Streams

TL;DR

MIDAS is an online, efficient algorithm for detecting microcluster anomalies in dynamic graph streams, offering theoretical guarantees and significantly outperforming existing methods in speed and accuracy.

Contribution

We introduce MIDAS, a novel microcluster-based anomaly detection method that operates in constant time and memory with proven false positive bounds.

Findings

Detects microcluster anomalies effectively in real-time.

Achieves 162-644 times faster processing than previous methods.

Improves detection accuracy by 42%-48% in AUC.

Abstract

Given a stream of graph edges from a dynamic graph, how can we assign anomaly scores to edges in an online manner, for the purpose of detecting unusual behavior, using constant time and memory? Existing approaches aim to detect individually surprising edges. In this work, we propose MIDAS, which focuses on detecting microcluster anomalies, or suddenly arriving groups of suspiciously similar edges, such as lockstep behavior, including denial of service attacks in network traffic data. MIDAS has the following properties: (a) it detects microcluster anomalies while providing theoretical guarantees about its false positive probability; (b) it is online, thus processing each edge in constant time and constant memory, and also processes the data 162-644 times faster than state-of-the-art approaches; (c) it provides 42%-48% higher accuracy (in terms of AUC) than state-of-the-art approaches.