Developing a generic framework for anomaly detection. (April 2022)

• Record Type: Journal Article
• Authors: Fatemifar, Soroush; Awais, Muhammad; Akbari, Ali; Kittler, Josef
• Is Part Of: Pattern recognition. Volume 124(2022)
• Publication Date: 2022-04
• Subjects: Anomaly detection -- One-class classification -- Score normalisation -- Face spoofing detection -- Convolutional neural network; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2021.108500
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: We develop a generic linear one-class classifier fusion method for anomaly detection that can be effectively applied to different domains and applications. We propose a new score normalisation method to support multiple classifier fusion, even in the case of the training data having a heavy-tailed distribution. We show that the performance of anomaly detection methods is normalisation sensitive. We define a novel fitness function to measure the effectiveness of the fused anomaly detectors without the need for anomalous samples and propose a particle swarm optimisation method for its optimisation. We experimentally and statistically demonstrate that the proposed weighted averaging fusion achieves superior performance compared to the state-of-the-art methods. Abstract: The fusion of one-class classifiers (OCCs) has been shown to exhibit promising performance in a variety of machine learning applications. The ability to assess the similarity or correlation between the output of various OCCs is an important prerequisite for building of a meaningful OCCs ensemble. However, this aspect of the OCC fusion problem has been mostly ignored so far. In this paper, we propose a new method of constructing a fusion of OCCs with three contributions: (a) As a key contribution, enabling an OCC ensemble design using exclusively non anomalous samples, we propose a novel fitness function to evaluate the competency of OCCs without requiring samples from the anomalous class; (b) As a minor, but impactful contribution, we investigate alternative forms of score normalisation of OCCs, and identify a novel two-sided normalisation method as the best in coping with long tail non anomalous data distributions; (c) In the context of building our proposed OCC fusion system based on the weighted averaging approach, we find that the weights optimised using a particle swarm optimisation algorithm produce the most effective solution. We evaluate the merits of the proposed method on 15 benchmarking datasets from different application domains including medical, anti-spam and face spoofing detection. The comparison of the proposed approach with state-of-the-art methods alongside the statistical analysis confirm the effectiveness of the proposed model.