Unsupervised Outlier Detection using Memory and Contrastive Learning

TL;DR

This paper introduces MCOD, a novel outlier detection framework using memory and contrastive learning to improve the discrimination of outliers from inliers in feature space, outperforming existing methods.

Contribution

The paper proposes a new outlier detection method combining memory modules and contrastive learning to enhance feature discrimination and detection accuracy.

Findings

MCOD outperforms nine state-of-the-art methods on four benchmark datasets.

Memory and contrastive modules improve feature space separation of outliers and inliers.

Extensive experiments validate the effectiveness of the proposed framework.

Abstract

Outlier detection is one of the most important processes taken to create good, reliable data in machine learning. The most methods of outlier detection leverage an auxiliary reconstruction task by assuming that outliers are more difficult to be recovered than normal samples (inliers). However, it is not always true, especially for auto-encoder (AE) based models. They may recover certain outliers even outliers are not in the training data, because they do not constrain the feature learning. Instead, we think outlier detection can be done in the feature space by measuring the feature distance between outliers and inliers. We then propose a framework, MCOD, using a memory module and a contrastive learning module. The memory module constrains the consistency of features, which represent the normal data. The contrastive learning module learns more discriminating features, which boosts the distinction between outliers and inliers. Extensive experiments on four benchmark datasets show that our proposed MCOD achieves a considerable performance and outperforms nine state-of-the-art methods.