Multivariate Time-Series Anomaly Detection with Contaminated Data

arXiv:2308.12563v2 Announce Type: replace
Abstract: Mainstream unsupervised anomaly detection algorithms often excel in academic datasets, yet their real-world performance is restricted due to the controlled experimental conditions involving clean training data. Addressing the challenge of training with noise, a prevalent issue in practical anomaly detection, is frequently overlooked. In a pioneering endeavor, this study delves into the realm of label-level noise within sensory time-series anomaly detection (TSAD). This paper presents a novel and practical end-to-end unsupervised TSAD when the training …

abstract academic algorithms anomaly anomaly detection arxiv challenge cs.lg data datasets detection eess.sp endeavor excel experimental issue multivariate noise performance practical series training training data type unsupervised world