# Automating Outlier Detection via Meta-Learning

arXiv: https://arxiv.org/abs/2009.10606 · HTML: https://mlsanity.com/paper/2009.10606/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 22 Sep 2020.

## Built on by (most cited first, 9 of 9)

- [A Comprehensive Survey on Graph Anomaly Detection with Deep Learning](https://mlsanity.com/paper/2106.07178/) (2021)
- [On the Detection of Reviewer-Author Collusion Rings From Paper Bidding](https://mlsanity.com/paper/2402.07860/) (2024)
- [A Meta-level Analysis of Online Anomaly Detectors](https://mlsanity.com/paper/2209.05899/) (2022)
- [Efficient Model Selection for Time Series Forecasting via LLMs](https://mlsanity.com/paper/2504.02119/) (2025)
- [We Need to Rethink Benchmarking in Anomaly Detection](https://mlsanity.com/paper/2507.15584/) (2025)
- [Anomaly Mining -- Past, Present and Future](https://mlsanity.com/paper/2105.10077/) (2021)
- [Towards Automated Self-Supervised Learning for Truly Unsupervised Graph Anomaly Detection](https://mlsanity.com/paper/2501.14694/) (2025)
- [UniOD: A Universal Model for Outlier Detection across Diverse Domains](https://mlsanity.com/paper/2507.06624/) (2025)
- [From Benchmarks to Production: Transferring Time Series Anomaly Detection Methods for Electricity Production Monitoring](https://mlsanity.com/paper/2609.39257/) (2026)

## Abstract (arXiv)

Given an unsupervised outlier detection (OD) task on a new dataset, how can we automatically select a good outlier detection method and its hyperparameter(s) (collectively called a model)? Thus far, model selection for OD has been a "black art"; as any model evaluation is infeasible due to the lack of (i) hold-out data with labels, and (ii) a universal objective function. In this work, we develop the first principled data-driven approach to model selection for OD, called MetaOD, based on meta-learning. MetaOD capitalizes on the past performances of a large body of detection models on existing outlier detection benchmark datasets, and carries over this prior experience to automatically select an effective model to be employed on a new dataset without using any labels. To capture task similarity, we introduce specialized meta-features that quantify outlying characteristics of a dataset. Through comprehensive experiments, we show the effectiveness of MetaOD in selecting a detection model that significantly outperforms the most popular outlier detectors (e.g., LOF and iForest) as well as various state-of-the-art unsupervised meta-learners while being extremely fast. To foster reproducibility and further research on this new problem, we open-source our entire meta-learning system, benchmark environment, and testbed datasets.