Automatic Unsupervised Ensemble Outlier Model Selection--Extended Version

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Phan, Hong-Phuc, Vu, Tuan-Anh, Kieu, Tung, Xuan, Son Ha, Yang, Bin, Jensen, Christian S.
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909049096765440
author Phan, Hong-Phuc
Vu, Tuan-Anh
Kieu, Tung
Xuan, Son Ha
Yang, Bin
Jensen, Christian S.
author_facet Phan, Hong-Phuc
Vu, Tuan-Anh
Kieu, Tung
Xuan, Son Ha
Yang, Bin
Jensen, Christian S.
contents Unsupervised outlier detection is attractive because it eliminates the need for labeled data. Moreover, forming multi-model ensembles can improve detection robustness. However, composing an ensemble without labeled data is challenging. Naively composed ensembles can suffer from ensemble saturation, where redundant or unreliable detection models degrade performance and incur unnecessary computation. We propose MetaEns, an automatic unsupervised framework for selecting ensembles of outlier detection models. Using labeled meta-datasets, MetaEns learns a model that predicts marginal ensemble gains, estimating the expected improvement from adding a candidate model to a partially constructed ensemble. At test time, this learned signal is combined with a submodular-inspired proxy objective that enforces diminishing returns through diversity-aware discounting and family-level risk regularization, thereby enabling greedy sequential selection with adaptive early stopping. As a result, MetaEns constructs compact, high-quality ensembles without access to ground-truth labels. Experiments on 39 real-world datasets show that MetaEns consistently outperforms state-of-the-art unsupervised selectors and ensemble baselines, achieving higher average precision while using fewer models.
format Preprint
id arxiv_https___arxiv_org_abs_2605_16567
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Automatic Unsupervised Ensemble Outlier Model Selection--Extended Version
Phan, Hong-Phuc
Vu, Tuan-Anh
Kieu, Tung
Xuan, Son Ha
Yang, Bin
Jensen, Christian S.
Machine Learning
Artificial Intelligence
Databases
Unsupervised outlier detection is attractive because it eliminates the need for labeled data. Moreover, forming multi-model ensembles can improve detection robustness. However, composing an ensemble without labeled data is challenging. Naively composed ensembles can suffer from ensemble saturation, where redundant or unreliable detection models degrade performance and incur unnecessary computation. We propose MetaEns, an automatic unsupervised framework for selecting ensembles of outlier detection models. Using labeled meta-datasets, MetaEns learns a model that predicts marginal ensemble gains, estimating the expected improvement from adding a candidate model to a partially constructed ensemble. At test time, this learned signal is combined with a submodular-inspired proxy objective that enforces diminishing returns through diversity-aware discounting and family-level risk regularization, thereby enabling greedy sequential selection with adaptive early stopping. As a result, MetaEns constructs compact, high-quality ensembles without access to ground-truth labels. Experiments on 39 real-world datasets show that MetaEns consistently outperforms state-of-the-art unsupervised selectors and ensemble baselines, achieving higher average precision while using fewer models.
title Automatic Unsupervised Ensemble Outlier Model Selection--Extended Version
topic Machine Learning
Artificial Intelligence
Databases
url https://arxiv.org/abs/2605.16567