Abstract
While the Dynamic Time Warping (DTW)-based Nearest-Neighbor Classification algorithm is regarded as a strong baseline for time series classification, in recent years there has been a plethora of algorithms that have claimed to be able to improve upon its accuracy in the general case. Many of these proposed ideas sacrifice the simplicity of implementation that DTW-based classifiers offer for rather modest gains. Nevertheless, there are clearly times when even a small improvement could make a large difference in an important medical or financial domain. In this work, we make an unexpected claim; an underappreciated 'low hanging fruit' in optimizing DTW's performance can produce improvements that make it an even stronger baseline, closing most or all the improvement gap of the more sophisticated methods. We show that the method currently used to learn DTW's only parameter, the maximum amount of warping allowed, is likely to give the wrong answer for small training sets. We introduce a simple method to mitigate the small training set issue by creating synthetic exemplars to help learn the parameter. We evaluate our ideas on the UCR Time Series Archive and a case study in fall classification, and demonstrate that our algorithm produces significant improvement in classification accuracy.
| Original language | English |
|---|---|
| Title of host publication | Proceedings - 2017 IEEE International Conference on Big Data |
| Subtitle of host publication | Dec 11- 14, 2017 • Boston, MA, USA |
| Editors | Jian-Yun Nie, Zoran Obradovic, Toyotaro Suzumura, Rumi Ghosh, Raghunath Nambiar, Chonggang Wang, Hui Zang, Ricardo Baeza-Yates, Xiaohua Hu, Jeremy Kepner, Alfredo Cuzzocrea, Jian Tang, Masashi Toyoda |
| Place of Publication | Piscataway NJ USA |
| Publisher | IEEE, Institute of Electrical and Electronics Engineers |
| Pages | 917-922 |
| Number of pages | 6 |
| ISBN (Electronic) | 9781538627143, 9781538627150 |
| ISBN (Print) | 9781538627167 |
| DOIs | |
| Publication status | Published - 2017 |
| Event | IEEE International Conference on Big Data (Big Data) 2017 - Boston, United States of America Duration: 11 Dec 2017 → 14 Dec 2017 Conference number: 5th http://cci.drexel.edu/bigdata/bigdata2017/ |
Conference
| Conference | IEEE International Conference on Big Data (Big Data) 2017 |
|---|---|
| Abbreviated title | Big Data 2017 |
| Country | United States of America |
| City | Boston |
| Period | 11/12/17 → 14/12/17 |
| Internet address |
Keywords
- classification
- Dynamic Time Warping
- time series
Cite this
}
Judicious setting of Dynamic Time Warping's window width allows more accurate classification of time series. / Dau, Hoang Anh; Silva, Diego Furtado; Petitjean, Francois; Forestier, Germain; Bagnall, Anthony; Keogh, Eamonn.
Proceedings - 2017 IEEE International Conference on Big Data: Dec 11- 14, 2017 • Boston, MA, USA. ed. / Jian-Yun Nie; Zoran Obradovic; Toyotaro Suzumura; Rumi Ghosh; Raghunath Nambiar; Chonggang Wang; Hui Zang; Ricardo Baeza-Yates; Xiaohua Hu; Jeremy Kepner; Alfredo Cuzzocrea; Jian Tang; Masashi Toyoda. Piscataway NJ USA : IEEE, Institute of Electrical and Electronics Engineers, 2017. p. 917-922.Research output: Chapter in Book/Report/Conference proceeding › Conference Paper › Research › peer-review
TY - GEN
T1 - Judicious setting of Dynamic Time Warping's window width allows more accurate classification of time series
AU - Dau, Hoang Anh
AU - Silva, Diego Furtado
AU - Petitjean, Francois
AU - Forestier, Germain
AU - Bagnall, Anthony
AU - Keogh, Eamonn
PY - 2017
Y1 - 2017
N2 - While the Dynamic Time Warping (DTW)-based Nearest-Neighbor Classification algorithm is regarded as a strong baseline for time series classification, in recent years there has been a plethora of algorithms that have claimed to be able to improve upon its accuracy in the general case. Many of these proposed ideas sacrifice the simplicity of implementation that DTW-based classifiers offer for rather modest gains. Nevertheless, there are clearly times when even a small improvement could make a large difference in an important medical or financial domain. In this work, we make an unexpected claim; an underappreciated 'low hanging fruit' in optimizing DTW's performance can produce improvements that make it an even stronger baseline, closing most or all the improvement gap of the more sophisticated methods. We show that the method currently used to learn DTW's only parameter, the maximum amount of warping allowed, is likely to give the wrong answer for small training sets. We introduce a simple method to mitigate the small training set issue by creating synthetic exemplars to help learn the parameter. We evaluate our ideas on the UCR Time Series Archive and a case study in fall classification, and demonstrate that our algorithm produces significant improvement in classification accuracy.
AB - While the Dynamic Time Warping (DTW)-based Nearest-Neighbor Classification algorithm is regarded as a strong baseline for time series classification, in recent years there has been a plethora of algorithms that have claimed to be able to improve upon its accuracy in the general case. Many of these proposed ideas sacrifice the simplicity of implementation that DTW-based classifiers offer for rather modest gains. Nevertheless, there are clearly times when even a small improvement could make a large difference in an important medical or financial domain. In this work, we make an unexpected claim; an underappreciated 'low hanging fruit' in optimizing DTW's performance can produce improvements that make it an even stronger baseline, closing most or all the improvement gap of the more sophisticated methods. We show that the method currently used to learn DTW's only parameter, the maximum amount of warping allowed, is likely to give the wrong answer for small training sets. We introduce a simple method to mitigate the small training set issue by creating synthetic exemplars to help learn the parameter. We evaluate our ideas on the UCR Time Series Archive and a case study in fall classification, and demonstrate that our algorithm produces significant improvement in classification accuracy.
KW - classification
KW - Dynamic Time Warping
KW - time series
UR - http://www.scopus.com/inward/record.url?scp=85047762839&partnerID=8YFLogxK
U2 - 10.1109/BigData.2017.8258009
DO - 10.1109/BigData.2017.8258009
M3 - Conference Paper
SN - 9781538627167
SP - 917
EP - 922
BT - Proceedings - 2017 IEEE International Conference on Big Data
A2 - Nie, Jian-Yun
A2 - Obradovic, Zoran
A2 - Suzumura, Toyotaro
A2 - Ghosh, Rumi
A2 - Nambiar, Raghunath
A2 - Wang, Chonggang
A2 - Zang, Hui
A2 - Baeza-Yates, Ricardo
A2 - Hu, Xiaohua
A2 - Kepner, Jeremy
A2 - Cuzzocrea, Alfredo
A2 - Tang, Jian
A2 - Toyoda, Masashi
PB - IEEE, Institute of Electrical and Electronics Engineers
CY - Piscataway NJ USA
ER -