Most methods for time series classification that attain state-of-the-art
accuracy have high computational complexity, requiring significant training
time even for smaller datasets, and are intractable for larger datasets.
Additionally, many existing methods focus on a single type of feature such as
shape or frequency. Building on the recent success of convolutional neural
networks for time series classification, we show that simple linear classifiers
using random convolutional kernels achieve state-of-the-art accuracy with a
fraction of the computational expense of existing methods.
%0 Generic
%1 dempster2019rocket
%A Dempster, Angus
%A Petitjean, François
%A Webb, Geoffrey I.
%D 2019
%K classification rocket timeseries todo:read
%T ROCKET: Exceptionally fast and accurate time series classification using
random convolutional kernels
%U http://arxiv.org/abs/1910.13051
%X Most methods for time series classification that attain state-of-the-art
accuracy have high computational complexity, requiring significant training
time even for smaller datasets, and are intractable for larger datasets.
Additionally, many existing methods focus on a single type of feature such as
shape or frequency. Building on the recent success of convolutional neural
networks for time series classification, we show that simple linear classifiers
using random convolutional kernels achieve state-of-the-art accuracy with a
fraction of the computational expense of existing methods.
@misc{dempster2019rocket,
abstract = {Most methods for time series classification that attain state-of-the-art
accuracy have high computational complexity, requiring significant training
time even for smaller datasets, and are intractable for larger datasets.
Additionally, many existing methods focus on a single type of feature such as
shape or frequency. Building on the recent success of convolutional neural
networks for time series classification, we show that simple linear classifiers
using random convolutional kernels achieve state-of-the-art accuracy with a
fraction of the computational expense of existing methods.},
added-at = {2021-04-28T08:39:27.000+0200},
author = {Dempster, Angus and Petitjean, François and Webb, Geoffrey I.},
biburl = {https://www.bibsonomy.org/bibtex/285b9dff8e978b7edbe8ade7558167e73/annakrause},
description = {1910.13051.pdf},
interhash = {a6f9cbdff453ca30e7b7b7efd1d7e772},
intrahash = {85b9dff8e978b7edbe8ade7558167e73},
keywords = {classification rocket timeseries todo:read},
note = {cite arxiv:1910.13051Comment: 27 pages, 23 figures},
timestamp = {2021-04-28T08:39:27.000+0200},
title = {ROCKET: Exceptionally fast and accurate time series classification using
random convolutional kernels},
url = {http://arxiv.org/abs/1910.13051},
year = 2019
}