Anomaly detection using spatial and temporal information in multivariate time series

Abstract

Real-world industrial systems contain a large number of interconnected sensors that generate a significant amount of time series data during system operation. Performing anomaly detection on these multivariate time series data can timely find faults, prevent malicious attacks, and ensure these systems safe and reliable operation. However, the rarity of abnormal instances leads to a lack of labeled data, so the supervised machine learning methods are not applicable. Furthermore, most current techniques do not take full advantage of the spatial and temporal dependencies implied among multiple variables to detect anomalies. Hence, we propose STADN, a novel Anomaly Detection Network Using Spatial and Temporal Information. STADN models the relationship graph between variables for a graph attention network to capture the spatial dependency between variables and utilizes a long short-term memory network to mine the temporal dependency of time series to fully use the spatial and temporal information of multivariate time series. STADN predicts the future behavior of each sensor by combining the historical behavior of the sensor and its neighbors, then detects and locates anomalies according to the prediction error. Furthermore, we improve the proposed model’s ability to discriminate anomaly and regularity and expand the prediction error gap between normal and abnormal instances by reconstructing the prediction errors. We conduct experiments on two real-world datasets, and the experimental results suggested that STADN achieves state-of-the-art outperformance.

AI-Assisted Fatigue and Stamina Control for Performance Sports on IMU-Generated Multivariate Times Series Datasets

“…where H(t) represents high-level features learned by the deep learning model (e.g., a Long Short-Term Memory (LSTM) network [48]), and Φ denotes the network parameters. We integrate an ensemble of models (E ) that captures different aspects of the data and learns diverse representations:…”

Accuracy Is Not Enough: Optimizing for a Fault Detection Delay

“…• Data: ANNs rely on sufficient high-quality data that accurately represent the problem. Data availability and quality can be a challenge, particularly because obtaining large, labeled datasets may be impractical or costly [4].…”

“…Although deep learning has shown success in fault detection, it is important to acknowledge that researchers using deep fault-detecting models sometimes overlook the crucial temporal aspect of fault detection. Specifically, they fail to leverage the existing temporal dependencies among variables using only a single data sample in one time step or, conversely, too much data in too many time steps [4]. Consequently, they fail to optimize for the time it takes to detect a fault.…”

“…Anomaly detection focuses on identifying data instances that deviate significantly from the norm [2]. In the context of anomaly detection for multivariate time-series, Tian identifies three major challenges [4]. First, the scarcity of labeled data arising from the rarity of outlier instances.…”

TimeTector: A Twin-Branch Approach for Unsupervised Anomaly Detection in Livestock Sensor Noisy Data (TT-TBAD)

“…Several approaches for analyzing spatial dependencies in multiple sensors have been studied [ 7 , 8 ]. To deal with such challenges, researchers have developed various methods, e.g., data-driven anomaly detection [ 9 ] and maximum mean discrepancy (MMD) [ 10 ]. Deep learning approaches, especially those involving autoencoders, have surfaced as potent solutions for outlier detection, although some may fall short in performance.…”