Abstract:Anomaly detection in multivariate time series data is of paramount importance for ensuring the efficient operation of large-scale systems across diverse domains. However, accurately detecting anomalies in such data poses significant challenges. Existing approaches, including forecasting and reconstruction-based methods, struggle to address these challenges effectively. To overcome these limitations, we propose a novel anomaly detection framework named ImDiffusion, which combines time series imputation and diffusion models to achieve accurate and robust anomaly detection. The imputation-based approach employed by ImDiffusion leverages the information from neighboring values in the time series, enabling precise modeling of temporal and inter-correlated dependencies, reducing uncertainty in the data, thereby enhancing the robustness of the anomaly detection process. ImDiffusion further leverages diffusion models as time series imputers to accurately capturing complex dependencies. We leverage the step-by-step denoised outputs generated during the inference process to serve as valuable signals for anomaly prediction, resulting in improved accuracy and robustness of the detection process. We evaluate the performance of ImDiffusion via extensive experiments on benchmark datasets. The results demonstrate that our proposed framework significantly outperforms state-of-the-art approaches in terms of detection accuracy and timeliness. ImDiffusion is further integrated into the real production system in Microsoft and observe a remarkable 11.4% increase in detection F1 score compared to the legacy approach. To the best of our knowledge, ImDiffusion represents a pioneering approach that combines imputation-based techniques with time series anomaly detection, while introducing the novel use of diffusion models to the field.

A new distributional treatment for time series anomaly detection

A New Distributional Treatment for Time Series and an Anomaly Detection Investigation.

Anomaly Detection at Scale: The Case for Deep Distributional Time Series Models

Anomaly Subsequence Detection with Dynamic Local Density for Time Series

Low-count Time Series Anomaly Detection

A novel unsupervised framework for time series data anomaly detection via spectrum decomposition

Clustering-based anomaly detection in multivariate time series data

Learning the feature distribution similarities for online time series anomaly detection

HybridAD: A Hybrid Model-Driven Anomaly Detection Approach for Multivariate Time Series

anomaly: Detection of Anomalous Structure in Time Series Data

Exact variable-length anomaly detection algorithm for univariate and multivariate time series

A Principled Distributional Approach to Trajectory Similarity Measurement and its Application to Anomaly Detection

Detecting Anomalies in Time Series Using Kernel Density Approaches

Local Subsequence-Based Distribution for Time Series Clustering

TFAD: A Decomposition Time Series Anomaly Detection Architecture with Time-Frequency Analysis

An Empirical Analysis of Anomaly Detection Methods for Multivariate Time Series

A space-embedding strategy for anomaly detection in multivariate time series

A simple method for unsupervised anomaly detection: An application to Web time series data

Anomaly Detection on Seasonal Metrics via Robust Time Series Decomposition

ImDiffusion: Imputed Diffusion Models for Multivariate Time Series Anomaly Detection

TimeSeriesBench: An Industrial-Grade Benchmark for Time Series Anomaly Detection Models