Abstract

In this paper, a discriminative autoencoding framework is proposed for semi-supervised anomaly detection using reconstruction errors. The framework only consists of a generator and a discriminative encoder, and the output of the latter is a vector. In the training process, the framework is trained as a generative adversarial network based on quadratic potential divergence. An extra loss added in the objective function enforces the discriminative encoder to use the mean value of the output vector for discrimination, which also empowers it with encoding ability. In the testing process, the trained framework can be used as an autoencoder to reconstruct test samples, where the trained discriminative encoder works as an encoder, and samples with reconstruction errors above a predefined threshold are determined as anomalies. The properties of quadratic potential divergence ensure a simple training process with comparable performance, meanwhile the discriminative encoder with two functions makes full use of training resources and reduces required network structures. Comparisons on benchmark datasets also show the efficiency and superiorities of the proposed methods.

Highlights

  • Anomaly detection refers to the problem of finding patterns in data that do not conform to the expected behaviors [1]

  • Detection has been applied to various fields including medical diagnostic problems [4], faults and failure detection in complex industrial systems [2], structural damage [5], intrusion detection [6], video surveillance [7], mobile robotics [8], and sensor networks [9]

  • This paper mainly focuses on the VAEs and Generative adversarial network (GAN) based frameworks for anomaly detection; on the other hand, the novel techniques can be applied in the summarized frameworks

Read more

Summary

Introduction

Detection refers to the problem of finding patterns in data that do not conform to the expected behaviors [1]. Detection has been applied to various fields including medical diagnostic problems [4], faults and failure detection in complex industrial systems [2], structural damage [5], intrusion detection [6], video surveillance [7], mobile robotics [8], and sensor networks [9]. With the increasing complexity of modern systems and surging volume of available data, it is more challenging to investigate the relationships among different components or data sources. Anomaly detection is playing an increasingly important role in data-related fields, and it is promising to investigate efficient solutions

Objectives
Methods
Findings
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call