Calibrated Unsupervised Anomaly Detection in Multivariate Time-series using Reinforcement Learning
This addresses the problem of high false negatives in anomaly detection for domains like industrial monitoring or healthcare, though it appears incremental as it builds on existing RL and autoencoder methods.
The paper tackles unsupervised anomaly detection in multivariate time-series by using reinforcement learning in an autoencoder's latent space, combined with wavelet analysis and synthetic anomaly calibration, achieving improved detection of both sudden and subtle anomalies.
This paper investigates unsupervised anomaly detection in multivariate time-series data using reinforcement learning (RL) in the latent space of an autoencoder. A significant challenge is the limited availability of anomalous data, often leading to misclassifying anomalies as normal events, thus raising false negatives. RL can help overcome this limitation by promoting exploration and balancing exploitation during training, effectively preventing overfitting. Wavelet analysis is also utilized to enhance anomaly detection, enabling time-series data decomposition into both time and frequency domains. This approach captures anomalies at multiple resolutions, with wavelet coefficients extracted to detect both sudden and subtle shifts in the data, thereby refining the anomaly detection process. We calibrate the decision boundary by generating synthetic anomalies and embedding a supervised framework within the model. This supervised element aids the unsupervised learning process by fine-tuning the decision boundary and increasing the model's capacity to distinguish between normal and anomalous patterns effectively.