Time Series Classification (TSC) is an important and challenging problem in
data mining. With the increase of time series data availability, hundreds of
TSC algorithms have been proposed. Among these methods, only a few have
considered Deep Neural Networks (DNNs) to perform this task. This is surprising
as deep learning has seen very successful applications in the last years. DNNs
have indeed revolutionized the field of computer vision especially with the
advent of novel deeper architectures such as Residual and Convolutional Neural
Networks. Apart from images, sequential data such as text and audio can also be
processed with DNNs to reach state-of-the-art performance for document
classification and speech recognition. In this article, we study the current
state-of-the-art performance of deep learning algorithms for TSC by presenting
an empirical study of the most recent DNN architectures for TSC. We give an
overview of the most successful deep learning applications in various time
series domains under a unified taxonomy of DNNs for TSC. We also provide an
open source deep learning framework to the TSC community where we implemented
each of the compared approaches and evaluated them on a univariate TSC
benchmark (the UCR/UEA archive) and 12 multivariate time series datasets. By
training 8,730 deep learning models on 97 time series datasets, we propose the
most exhaustive study of DNNs for TSC to date.
Description
Deep learning for time series classification: a review - 1809.04356.pdf
%0 Generic
%1 fawaz2018learning
%A Fawaz, Hassan Ismail
%A Forestier, Germain
%A Weber, Jonathan
%A Idoumghar, Lhassane
%A Muller, Pierre-Alain
%D 2018
%K classification multivariate survey timeSeries todo:read
%R 10.1007/s10618-019-00619-1
%T Deep learning for time series classification: a review
%U http://arxiv.org/abs/1809.04356
%X Time Series Classification (TSC) is an important and challenging problem in
data mining. With the increase of time series data availability, hundreds of
TSC algorithms have been proposed. Among these methods, only a few have
considered Deep Neural Networks (DNNs) to perform this task. This is surprising
as deep learning has seen very successful applications in the last years. DNNs
have indeed revolutionized the field of computer vision especially with the
advent of novel deeper architectures such as Residual and Convolutional Neural
Networks. Apart from images, sequential data such as text and audio can also be
processed with DNNs to reach state-of-the-art performance for document
classification and speech recognition. In this article, we study the current
state-of-the-art performance of deep learning algorithms for TSC by presenting
an empirical study of the most recent DNN architectures for TSC. We give an
overview of the most successful deep learning applications in various time
series domains under a unified taxonomy of DNNs for TSC. We also provide an
open source deep learning framework to the TSC community where we implemented
each of the compared approaches and evaluated them on a univariate TSC
benchmark (the UCR/UEA archive) and 12 multivariate time series datasets. By
training 8,730 deep learning models on 97 time series datasets, we propose the
most exhaustive study of DNNs for TSC to date.
@misc{fawaz2018learning,
abstract = {Time Series Classification (TSC) is an important and challenging problem in
data mining. With the increase of time series data availability, hundreds of
TSC algorithms have been proposed. Among these methods, only a few have
considered Deep Neural Networks (DNNs) to perform this task. This is surprising
as deep learning has seen very successful applications in the last years. DNNs
have indeed revolutionized the field of computer vision especially with the
advent of novel deeper architectures such as Residual and Convolutional Neural
Networks. Apart from images, sequential data such as text and audio can also be
processed with DNNs to reach state-of-the-art performance for document
classification and speech recognition. In this article, we study the current
state-of-the-art performance of deep learning algorithms for TSC by presenting
an empirical study of the most recent DNN architectures for TSC. We give an
overview of the most successful deep learning applications in various time
series domains under a unified taxonomy of DNNs for TSC. We also provide an
open source deep learning framework to the TSC community where we implemented
each of the compared approaches and evaluated them on a univariate TSC
benchmark (the UCR/UEA archive) and 12 multivariate time series datasets. By
training 8,730 deep learning models on 97 time series datasets, we propose the
most exhaustive study of DNNs for TSC to date.},
added-at = {2021-04-28T08:28:37.000+0200},
author = {Fawaz, Hassan Ismail and Forestier, Germain and Weber, Jonathan and Idoumghar, Lhassane and Muller, Pierre-Alain},
biburl = {https://www.bibsonomy.org/bibtex/25ac7523bd04bd83771e2356e7e9e6089/annakrause},
description = {Deep learning for time series classification: a review - 1809.04356.pdf},
doi = {10.1007/s10618-019-00619-1},
interhash = {dda33123411f47e9d56ec2357eaaaec6},
intrahash = {5ac7523bd04bd83771e2356e7e9e6089},
keywords = {classification multivariate survey timeSeries todo:read},
note = {cite arxiv:1809.04356Comment: Accepted at Data Mining and Knowledge Discovery},
timestamp = {2022-09-13T17:20:58.000+0200},
title = {Deep learning for time series classification: a review},
url = {http://arxiv.org/abs/1809.04356},
year = 2018
}