Deep anomaly detection using self-supervised learning: application to time series of cellular data - Archive ouverte HAL Access content directly
Conference Papers Year : 2021

Deep anomaly detection using self-supervised learning: application to time series of cellular data

(1, 2) , (1) , (3) , (4) , (2)
1
2
3
4

Abstract

We present a deep self-supervised method for anomaly detection on time series. We apply this methodology to detect anomalies from cellular times series, in particular cell dry mass, obtained in the context of lens-free microscopy.We propose an innovative, self-supervised, two-step method for anomaly detection on time series. As a first step, a representation of the time series is learned thanks to a 1D-convolutional neural network without any labels. Then, the learnedrepresentation is used to feed a threshold anomaly detector. This new self-supervised learning method is tested on an unlabeleddataset of 9100 time series of dry mass and succeeded in detecting abnormal time series with a precision of 96.6%.
Fichier principal
Vignette du fichier
ASPAI21.pdf (612.87 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

cea-03605065 , version 1 (24-11-2021)
cea-03605065 , version 2 (10-03-2022)

Identifiers

  • HAL Id : cea-03605065 , version 2

Cite

Romain Bailly, Marielle Malfante, Cédric Allier, Lamya Ghenim, Jérôme I. Mars. Deep anomaly detection using self-supervised learning: application to time series of cellular data. ASPAI 2021 - 3rd International Conference on Advances in Signal Processing and Artificial Intelligence, Nov 2021, Porto, Portugal. ⟨cea-03605065v2⟩
222 View
316 Download

Share

Gmail Facebook Twitter LinkedIn More