Tighten The Lasso: A Convex Hull Volume-based Anomaly Detection Method

TL;DR

This paper introduces a convex hull volume-based anomaly detection method that identifies out-of-distribution data by monitoring volume changes, offering a novel approach with competitive performance and efficiency advantages.

Contribution

The paper presents a new convex hull volume-based algorithm for OOD detection, exploiting volume changes to distinguish in-distribution from OOD data, and provides an efficient criterion for dataset suitability.

Findings

Performs comparably to state-of-the-art methods across multiple datasets.

Identifies datasets where the method outperforms existing approaches.

Offers a computationally efficient criterion for method applicability.

Abstract

Detecting out-of-distribution (OOD) data is a critical task for maintaining model reliability and robustness. In this study, we propose a novel anomaly detection algorithm that leverages the convex hull (CH) property of a dataset by exploiting the observation that OOD samples marginally increase the CH's volume compared to in-distribution samples. Thus, we establish a decision boundary between OOD and in-distribution data by iteratively computing the CH's volume as samples are removed, stopping when such removal does not significantly alter the CH's volume. The proposed algorithm is evaluated against seven widely used anomaly detection methods across ten datasets, demonstrating performance comparable to state-of-the-art (SOTA) techniques. Furthermore, we introduce a computationally efficient criterion for identifying datasets where the proposed method outperforms existing SOTA approaches.