F4: An All-Purpose Tool for Multivariate Time Series Classification
We propose Fast Forest of Flexible Features (F4), a novel approach for classifying multivariate time series, which is aimed to discriminate between underlying generating processes. This goal has barely been addressed in the literature. F4 consists of two steps. First, a set of features based on the...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-11-01
|
| Series: | Mathematics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7390/9/23/3051 |
| _version_ | 1797507412495695872 |
|---|---|
| author | Ángel López-Oriona José A. Vilar |
| author_facet | Ángel López-Oriona José A. Vilar |
| author_sort | Ángel López-Oriona |
| collection | DOAJ |
| description | We propose Fast Forest of Flexible Features (F4), a novel approach for classifying multivariate time series, which is aimed to discriminate between underlying generating processes. This goal has barely been addressed in the literature. F4 consists of two steps. First, a set of features based on the quantile cross-spectral density and the maximum overlap discrete wavelet transform are extracted from each series. Second, a random forest is fed with the extracted features. An extensive simulation study shows that F4 outperforms some powerful classifiers in a wide variety of situations, including stationary and nonstationary series. The proposed method is also capable of successfully discriminating between electrocardiogram (ECG) signals of healthy subjects and those with myocardial infarction condition. Additionally, despite lacking shape-based information, F4 attains state-of-the-art results in some datasets of the University of East Anglia (UEA) multivariate time series classification archive. |
| first_indexed | 2024-03-10T04:48:07Z |
| format | Article |
| id | doaj.art-2a038fa6c179424a9867340d7037a98e |
| institution | Directory Open Access Journal |
| issn | 2227-7390 |
| language | English |
| last_indexed | 2024-03-10T04:48:07Z |
| publishDate | 2021-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Mathematics |
| spelling | doaj.art-2a038fa6c179424a9867340d7037a98e2023-11-23T02:45:17ZengMDPI AGMathematics2227-73902021-11-01923305110.3390/math9233051F4: An All-Purpose Tool for Multivariate Time Series ClassificationÁngel López-Oriona0José A. Vilar1Research Group MODES, Research Center for Information and Communication Technologies (CITIC), University of A Coruña, 15071 A Coruña, SpainResearch Group MODES, Research Center for Information and Communication Technologies (CITIC), University of A Coruña, 15071 A Coruña, SpainWe propose Fast Forest of Flexible Features (F4), a novel approach for classifying multivariate time series, which is aimed to discriminate between underlying generating processes. This goal has barely been addressed in the literature. F4 consists of two steps. First, a set of features based on the quantile cross-spectral density and the maximum overlap discrete wavelet transform are extracted from each series. Second, a random forest is fed with the extracted features. An extensive simulation study shows that F4 outperforms some powerful classifiers in a wide variety of situations, including stationary and nonstationary series. The proposed method is also capable of successfully discriminating between electrocardiogram (ECG) signals of healthy subjects and those with myocardial infarction condition. Additionally, despite lacking shape-based information, F4 attains state-of-the-art results in some datasets of the University of East Anglia (UEA) multivariate time series classification archive.https://www.mdpi.com/2227-7390/9/23/3051multivariate time seriesclassificationquantile analysiswavelet analysisrandom forestECG signals |
| spellingShingle | Ángel López-Oriona José A. Vilar F4: An All-Purpose Tool for Multivariate Time Series Classification Mathematics multivariate time series classification quantile analysis wavelet analysis random forest ECG signals |
| title | F4: An All-Purpose Tool for Multivariate Time Series Classification |
| title_full | F4: An All-Purpose Tool for Multivariate Time Series Classification |
| title_fullStr | F4: An All-Purpose Tool for Multivariate Time Series Classification |
| title_full_unstemmed | F4: An All-Purpose Tool for Multivariate Time Series Classification |
| title_short | F4: An All-Purpose Tool for Multivariate Time Series Classification |
| title_sort | f4 an all purpose tool for multivariate time series classification |
| topic | multivariate time series classification quantile analysis wavelet analysis random forest ECG signals |
| url | https://www.mdpi.com/2227-7390/9/23/3051 |
| work_keys_str_mv | AT angellopezoriona f4anallpurposetoolformultivariatetimeseriesclassification AT joseavilar f4anallpurposetoolformultivariatetimeseriesclassification |