# Interpretable Maximum Margin Deep Anomaly Detection

arXiv: https://arxiv.org/abs/2603.07073 · HTML: https://mlsanity.com/paper/2603.07073/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 12 of its 33 references that are in the map.
- First posted on arXiv 7 Mar 2026.

## Builds on (most cited first, 12 of 12)

- [Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization](https://mlsanity.com/paper/1610.02391/) (2016)
- [Learning Deep Features for Discriminative Localization](https://mlsanity.com/paper/1512.04150/) (2015)
- [A Survey on Neural Network Interpretability](https://mlsanity.com/paper/2012.14261/) (2020)
- [Explainable Deep One-Class Classification](https://mlsanity.com/paper/2007.01760/) (2020)
- [DROCC: Deep Robust One-Class Classification](https://mlsanity.com/paper/2002.12718/) (2020)
- [ReContrast: Domain-Specific Anomaly Detection via Contrastive Reconstruction](https://mlsanity.com/paper/2306.02602/) (2023)
- [MOCCA: Multi-Layer One-Class ClassificAtion for Anomaly Detection](https://mlsanity.com/paper/2012.12111/) (2020)
- [Towards Explaining Anomalies: A Deep Taylor Decomposition of One-Class Models](https://mlsanity.com/paper/1805.06230/) (2018)
- [Perturbation Learning Based Anomaly Detection](https://mlsanity.com/paper/2206.02704/) (2022)
- [DASVDD: Deep Autoencoding Support Vector Data Descriptor for Anomaly Detection](https://mlsanity.com/paper/2106.05410/) (2021)
- [Deep Orthogonal Hypersphere Compression for Anomaly Detection](https://mlsanity.com/paper/2302.06430/) (2023)
- [Deep Dual Support Vector Data Description for Anomaly Detection on Attributed Networks](https://mlsanity.com/paper/2109.00138/) (2021)

## Built on by (most cited first, 2 of 2)

- [Deep Convolutional Large-Margin ℓp-SVDD for Visual Anomaly Detection](https://mlsanity.com/paper/2607.22212/) (2026)
- [Interpretable Multi-Hypersphere Deep Anomaly Detection for Open-set Supervised Anomaly Detection](https://mlsanity.com/paper/2609.23008/) (2026)

## Abstract (arXiv)

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.