Coverage-based Outlier Explanation

TL;DR

This paper introduces a new method for explaining outliers in datasets by identifying interpretable features that characterize outliers, aiding domain experts in understanding anomalies.

Contribution

It formulates outlier explanation as an optimization problem focused on interpretability and purity, advancing beyond traditional detection algorithms.

Findings

Efficiently generates better explanations than rule-based learners.

Provides semantic explanations understandable to domain experts.

Demonstrates effectiveness on real-world datasets.

Abstract

Outlier detection is a core task in data mining with a plethora of algorithms that have enjoyed wide scale usage. Existing algorithms are primarily focused on detection, that is the identification of outliers in a given dataset. In this paper we explore the relatively under-studied problem of the outlier explanation problem. Our goal is, given a dataset that is already divided into outliers and normal instances, explain what characterizes the outliers. We explore the novel direction of a semantic explanation that a domain expert or policy maker is able to understand. We formulate this as an optimization problem to find explanations that are both interpretable and pure. Through experiments on real-world data sets, we quantitatively show that our method can efficiently generate better explanations compared with rule-based learners.