Lune

ICLR2025Top-tier venue

AutoUAD: Hyper-parameter Optimization for Unsupervised Anomaly Detection

Wei Dai, Jicong Fan

Abstract

Unsupervised anomaly detection (UAD) has important applications in diverse fields such as manufacturing industry and medical diagnosis. In the past decades, although numerous insightful and effective UAD methods have been proposed, it remains a huge challenge to tune the hyper-parameters of each method and select the most appropriate method among many candidates for a specific dataset, due to the absence of labeled anomalies in the training phase of UAD methods and the high diversity of real datasets. In this work, we aim to address this challenge, so as to make UAD more practical and reliable. We propose two internal evaluation metrics, relative-top-median and expected-anomaly-gap, and one semi-internal evaluation metric, normalized pseudo discrepancy (NPD), as surrogate functions of the expected model performance on unseen test data. For instance, NPD measures the discrepancy between the anomaly scores of a validation set drawn from the training data and a validation set drawn from an isotropic Gaussian. NPD is simple and hyper-parameter-free and is able to compare different UAD methods, and its effectiveness is theoretically analyzed. We integrate the three metrics with Bayesian optimization to effectively optimize the hyper-parameters of UAD models. Extensive experiments on 38 datasets show the effectiveness of our methods.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

4 top-tier papers have cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • Self-Discriminative Modeling for Anomalous Graph Detection cites Neural Transformation Learning for Deep Anomaly Detection Beyond Images
  • Fairness-aware Anomaly Detection via Fair Projection cites Self-Discriminative Modeling for Anomalous Graph Detection
  • Fairness-aware Anomaly Detection via Fair Projection cites DROCC: Deep Robust One-Class Classification
  • Neural Transformation Learning for Deep Anomaly Detection Beyond Images cites DROCC: Deep Robust One-Class Classification
  • UniOD: A Universal Model for Outlier Detection across Diverse Domains cites Self-Discriminative Modeling for Anomalous Graph Detection
  • UniOD: A Universal Model for Outlier Detection across Diverse Domains cites Fairness-aware Anomaly Detection via Fair Projection
  • UniOD: A Universal Model for Outlier Detection across Diverse Domains cites Neural Transformation Learning for Deep Anomaly Detection Beyond Images
  • UniOD: A Universal Model for Outlier Detection across Diverse Domains cites Anomaly Detection for Tabular Data with Internal Contrastive Learning
  • Noise-Robust Density Estimation for Tabular Data Anomaly Detection cites Neural Transformation Learning for Deep Anomaly Detection Beyond Images
  • Noise-Robust Density Estimation for Tabular Data Anomaly Detection cites UniOD: A Universal Model for Outlier Detection across Diverse Domains
  • Noise-Robust Density Estimation for Tabular Data Anomaly Detection cites Anomaly Detection for Tabular Data with Internal Contrastive Learning
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers4

  1. Fairness-aware Anomaly Detection via Fair ProjectionFeng Xiao, Xiaoying Tang, Jicong FanNeurIPS 2025 · 2 citations
  2. UniOD: A Universal Model for Outlier Detection across Diverse DomainsDazhi Fu, Jicong FanICLR 2026 · 1 citation
  3. Self-Discriminative Modeling for Anomalous Graph DetectionJinyu Cai, Yunhe Zhang, Jicong FanICML 2025
  4. Noise-Robust Density Estimation for Tabular Data Anomaly DetectionDazhi Fu, Zhao Zhang, Jicong FanICML 2026

Builds on16

  1. Out-of-Distribution Detection with Deep Nearest NeighborsYiyou Sun, Yifei Ming, Xiaojin Zhu, Yixuan LiICML 2022 · 789 citations
  2. Learning Semantic Context from Normal Samples for Unsupervised Anomaly DetectionXudong Yan, Huaidong Zhang, Xuemiao Xu, Xiaowei Hu et al.AAAI 2021 · 210 citations
  3. DROCC: Deep Robust One-Class ClassificationSachin Goyal, Aditi Raghunathan, Moksh Jain, Harsha Vardhan Simhadri et al.ICML 2020 · 202 citations
  4. Neural Transformation Learning for Deep Anomaly Detection Beyond ImagesChen Qiu, Timo Pfrommer, Marius Kloft, Stephan Mandt et al.ICML 2021 · 171 citations
  5. Anomaly Detection for Tabular Data with Internal Contrastive LearningTom Shenkar, Lior WolfICLR 2022 · 127 citations

Related papers

  1. Unsupervised Model Selection for Time Series Anomaly DetectionMononito Goswami, Cristian I. Challu, Laurent Callot, Lenon Minorics et al.ICLR 2023 · 8 citations
  2. TSB-UAD: An End-to-End Benchmark Suite for Univariate Time-Series Anomaly DetectionJohn Paparrizos, Yuhao Kang, Paul Boniol, Ruey S. Tsay et al.VLDB 2022 · 138 citations
  3. TSB-AutoAD: Towards Automated Solutions for Time-Series Anomaly Detection [E, A & B]Qinghua Liu, Seunghak Lee, John PaparrizosVLDB 2025 · 13 citations
  4. SEAD: Unsupervised Ensemble of Streaming Anomaly DetectorsSaumya Gaurang Shah, Abishek Sankararaman, Balakrishnan Narayanaswamy, Vikramank Y. SinghICML 2025
  5. Volume Under the Surface: A New Accuracy Evaluation Measure for Time-Series Anomaly DetectionJohn Paparrizos, Paul Boniol, Themis Palpanas, Ruey S. Tsay et al.VLDB 2022 · 171 citations
Dusk over the sea between two cliffs drawn in fine vertical lines