Anomaly Detection with Selective Dictionary Learning

TL;DR

This paper introduces novel unsupervised anomaly detection methods using Dictionary Learning and Kernel Dictionary Learning, including a reduced kernel version for large datasets and a random signal selection technique to improve robustness.

Contribution

It adapts existing DL and KDL algorithms for anomaly detection, introduces a reduced kernel version (RKDL), and enhances methods with random signal selection to eliminate outliers during training.

Findings

Algorithms outperform standard benchmarks

Reduced kernel version handles large datasets efficiently

Random selection improves outlier elimination

Abstract

In this paper we present new methods of anomaly detection based on Dictionary Learning (DL) and Kernel Dictionary Learning (KDL). The main contribution consists in the adaption of known DL and KDL algorithms in the form of unsupervised methods, used for outlier detection. We propose a reduced kernel version (RKDL), which is useful for problems with large data sets, due to the large kernel matrix. We also improve the DL and RKDL methods by the use of a random selection of signals, which aims to eliminate the outliers from the training procedure. All our algorithms are introduced in an anomaly detection toolbox and are compared to standard benchmark results.