Interpretable Maximum Margin Deep Anomaly Detection

2026-03-07

AI Summary

This work addresses key challenges in deep one-class anomaly detection—namely, hypersphere collapse, heuristic reliance on hypersphere parameter initialization, and limited model interpretability—by introducing an end-to-end trainable framework that leverages a small number of labeled anomalous samples. The proposed method equivalently represents the hypersphere center and radius as the weights of the network’s final layer and jointly optimizes these parameters through a max-margin objective, inherently preventing collapse. This formulation enables efficient training while offering intrinsic interpretability: the learned weights can be directly visualized to reveal the model’s decision rationale. Experimental results demonstrate that the approach significantly outperforms state-of-the-art methods on both image and tabular benchmarks, while simultaneously producing interpretable outputs of diagnostic value.

Technology Category

• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification
• Computer Vision: Interpretability, Explainability, and Transparency
• Data Mining & Knowledge Management: Anomaly/Outlier Detection

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Responsible Web: Algorithmic accountability and transparency on the web

Abstract

Anomaly detection is a crucial machine-learning task with wide-ranging applications. Deep Support Vector Data Description (Deep SVDD) is a prominent deep one-class method, but it is vulnerable to hypersphere collapse, often relies on heuristic choices for hypersphere parameters, and provides limited interpretability. To address these issues, we propose Interpretable Maximum Margin Deep Anomaly Detection (IMD-AD), which leverages a small set of labeled anomalies and a maximum margin objective to stabilize training and improve discrimination. It is inherently resilient to hypersphere collapse. Furthermore, we prove an equivalence between hypersphere parameters and the network's final-layer weights, which allows the center and radius to be learned end-to-end as part of the model and yields intrinsic interpretability and visualizable outputs. We further develop an efficient training algorithm that jointly optimizes representation, margin, and final-layer parameters. Extensive experiments and ablation studies on image and tabular benchmarks demonstrate that IMD-AD empirically improves detection performance over several state-of-the-art baselines while providing interpretable decision diagnostics.

Problem

• anomaly detection
• hypersphere collapse
• interpretability
• deep one-class classification
• maximum margin

Innovation

• maximum margin
• hypersphere collapse
• end-to-end learning
• interpretability
• deep anomaly detection