Personalized anomaly detection using deep active learning

RAS Techniques and Instruments, vol. 2, pp. 586–598

Abstract

Abstract Anomaly detection algorithms are typically applied to static, unchanging, data features hand-crafted by the user. But how does a user systematically craft good features for anomalies that have never been seen? Here we couple deep learning with active learning – in which an Oracle iteratively labels small amounts of data selected algorithmically over a series of rounds – to automatically and dynamically improve the data features for efficient outlier detection. This approach, Anomaly Hunt (ahunt), shows excellent performance on MNIST, CIFAR10, and Galaxy-DECaLS data, significantly outperforming both standard anomaly detection and active learning algorithms with static feature spaces. Beyond improved performance, ahunt also allows the number of anomaly classes to grow organically in response to the Oracle’s evaluations. Extensive ablation studies explore the impact of Oracle question selection strategy and loss function on performance. We illustrate how the dynamic anomaly class taxonomy represents another step towards fully personalized rankings of different anomaly classes that reflect a user’s interests, allowing the algorithm to learn to ignore statistically significant but uninteresting outliers (e.g. noise). This should prove useful in the era of massive astronomical data sets serving diverse sets of users who can only review a tiny subset of the incoming data.

Authors 3

1. Alireza Vafaei Sadr corresponding Aachen Institute of Pathology

   RWTH Aachen University · Universitätsklinikum Aachen · University of Geneva · Pennsylvania State University

   Affiliation as printed

   Department of Public Health Sciences, College of Medicine, The Pennsylvania State University , Hershey, PA 17033 , USA

   Département de Physique Théorique and Center for Astroparticle Physics, University of Geneva , 1211 Geneva , Switzerland

   Institute of Pathology, University Hospital RWTH Aachen , Pauwelsstraße 30, D-52074, Aachen , Germany

2. Bruce A. Bassett

   African Institute for Mathematical Sciences · University of Cape Town · South African Astronomical Observatory

   Affiliation as printed

   African Institute for Mathematical Sciences , 6 Melrose Road, Muizenberg 7945 , South Africa

   Department of Maths and Applied Maths, University of Cape Town , Cape Town, 7700 , South Africa

   South African Astronomical Observatory , Cape Town 7925 , South Africa

   South African Astronomical Observatory, Observatory, Cape Town 7925, South Africa

3. Emmanuel Sekyi

   African Institute for Mathematical Sciences · University of Cape Town

   Affiliation as printed

   African Institute for Mathematical Sciences , 6 Melrose Road, Muizenberg 7945 , South Africa

   Department of Maths and Applied Maths, University of Cape Town , Cape Town, 7700 , South Africa