Abstract:Developing efficient time series anomaly detection techniques is important to maintain service quality and provide early alarms. Generative neural network methods are one class of the unsupervised approaches that are achieving increasing attention in recent years. In this paper, we have proposed an unsupervised Generative Adversarial Network (GAN)-based anomaly detection framework, DEGAN. It relies solely on normal time series data as input to train a well-configured discriminator (D) into a standalone anomaly predictor. In this framework, time series data is processed by the sliding window method. Expected normal patterns in data are leveraged to develop a generator (G) capable of generating normal data patterns. Normal data is also utilized in hyperparameter tuning and D model selection steps. Validated D models are then extracted and applied to evaluate unseen (testing) time series and identify patterns that have anomalous characteristics. Kernel density estimation (KDE) is applied to data points that are likely to be anomalous to generate probability density functions on the testing time series. The segments with the highest relative probabilities are detected as anomalies. To evaluate the performance, we tested on univariate acceleration time series for five miles of a Class I railroad track. We implemented the framework to detect the real anomalous observations identified by operators. The results show that leveraging the framework with a CNN D architecture results in average best recall and precision of 80% and 86%, respectively, which demonstrates that a well-trained standalone D model has the potential to be a reliable anomaly detector. Moreover, the influence of GAN hyperparameters, GAN architectures, sliding window sizes, clustering of time series, and model validation with labeled/unlabeled data were also investigated.

STAD-GAN: Unsupervised Anomaly Detection on Multivariate Time Series with Self-training Generative Adversarial Networks

Data anomaly detection for structural health monitoring using a combination network of GANomaly and CNN

TSI-GAN: Unsupervised Time Series Anomaly Detection using Convolutional Cycle-Consistent Generative Adversarial Networks

Robust Unsupervised Anomaly Detection Via Multi-Time Scale DCGANs with Forgetting Mechanism for Industrial Multivariate Time Series

MTS-DVGAN: Anomaly Detection in Cyber-Physical Systems using a Dual Variational Generative Adversarial Network

A Novel Multi-dimensional Time-series Data Anomaly Detection Model Based on Generative Adversarial Network Aided Autoencoder

Self-Supervised Spatial-Temporal Normality Learning for Time Series Anomaly Detection

Research on DUAL-ADGAN Model for Anomaly Detection Method in Time-Series Data

Multi-scale Anomaly Detection for Big Time Series of Industrial Sensors

GTAD: Graph and Temporal Neural Network for Multivariate Time Series Anomaly Detection

DEGAN: Time Series Anomaly Detection using Generative Adversarial Network Discriminators and Density Estimation

Attention Based CNN-LSTM Network for Anomaly Pattern Classification of Multivariate Time Series

Unsupervised Anomaly Detection Via Generative Adversarial Networks: Poster Abstract

Management Analysis Method of Multivariate Time Series Anomaly Detection in Financial Risk Assessment

Anomaly detection by using a combination of generative adversarial networks and convolutional autoencoders

MST-GAT: A Multimodal Spatial-Temporal Graph Attention Network for Time Series Anomaly Detection

Supervised Anomaly Detection via Conditional Generative Adversarial Network and Ensemble Active Learning

GANomaly: Semi-supervised Anomaly Detection via Adversarial Training

GAN-based Data Augmentation Strategy for Sensor Anomaly Detection in Industrial Robots

MIM-GAN-based Anomaly Detection for Multivariate Time Series Data

Adversarially Learned Anomaly Detection