Time Series Anomaly Detection using Diffusion-based Models

Ioana Pintilie, Andrei Manolache, Florin Brad
11/2/2023
cs.LGcs.AI

Official code: fbrad/DiffusionAE (40 stars, last commit 2023-11-20).

Abstract

Diffusion models have been recently used for anomaly detection (AD) in images. In this paper we investigate whether they can also be leveraged for AD on multivariate time series (MTS). We test two diffusion-based models and compare them to several strong neural baselines. We also extend the PA%K protocol, by computing a ROCK-AUC metric, which is agnostic to both the detection threshold and the ratio K of correctly detected points. Our models outperform the baselines on synthetic datasets and are competitive on real-world datasets, illustrating the potential of diffusion-based methods for AD in multivariate time series.

View on arXivView PDF

Code Implementations(1)

fbrad/DiffusionAEOfficial100%
406Python, Jupyter NotebookApr 3, 20232 years ago

Similar Papers

Physics-Informed Diffusion Models for Unsupervised Anomaly Detection in Multivariate Time Series

90% match2025

Multivariate Time Series Anomaly Detection using DiffGAN Model

89% match1 repo2025

On Diffusion Modeling for Anomaly Detection

89% match1 repo472023

Research on Anomaly Detection Methods Based on Diffusion Models

88% match2025

Anomaly detection using Diffusion-based methods

88% match2024

ImDiffusion: Imputed Diffusion Models for Multivariate Time Series Anomaly Detection

88% match1 repo1162023

Cite this paper

@article{pintilie2023time,
  title  = {Time Series Anomaly Detection using Diffusion-based Models},
  author = {Ioana Pintilie and Andrei Manolache and Florin Brad},
  year   = {2023},
  eprint = {2311.01452},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2311.01452v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2311.01452.svg)](https://codeofpaper.com/paper/2311.01452)

Discussion