AMP: a new time-frequency feature extraction method for intermittent time-series data

The characterisation of time-series data via their most salient features is extremely important in a range of machine learning task, not least of all with regards to classification and clustering. While there exist many feature extraction techniques suitable for non-intermittent time-series data, th...

Full description

Bibliographic Details
Main Authors: Barrack, Duncan S., Goulding, James, Hopcraft, Keith, Preston, Simon, Smith, Gavin
Format: Conference or Workshop Item
Published: 2015
Subjects:
Online Access:https://eprints.nottingham.ac.uk/52186/
_version_ 1848798668306513920
author Barrack, Duncan S.
Goulding, James
Hopcraft, Keith
Preston, Simon
Smith, Gavin
author_facet Barrack, Duncan S.
Goulding, James
Hopcraft, Keith
Preston, Simon
Smith, Gavin
author_sort Barrack, Duncan S.
building Nottingham Research Data Repository
collection Online Access
description The characterisation of time-series data via their most salient features is extremely important in a range of machine learning task, not least of all with regards to classification and clustering. While there exist many feature extraction techniques suitable for non-intermittent time-series data, these approaches are not always appropriate for intermittent time-series data, where intermittency is characterized by constant values for large periods of time punctuated by sharp and transient increases or decreases in value. Motivated by this, we present aggregation, mode decomposition and projection (AMP) a feature extraction technique particularly suited to intermittent time-series data which contain time-frequency patterns. For our method all individual time-series within a set are combined to form a non-intermittent aggregate. This is decomposed into a set of components which represent the intrinsic time-frequency signals within the data set. Individual time-series can then be _t to these components to obtain a set of numerical features that represent their intrinsic time-frequency patterns. To demonstrate the effectiveness of AMP, we evaluate against the real word task of clustering intermittent time-series data. Using synthetically generated data we show that a clustering approach which uses the features derived from AMP significantly outperforms traditional clustering methods. Our technique is further exemplified on a real world data set where AMP can be used to discover groupings of individuals which correspond to real world sub-populations.
first_indexed 2025-11-14T20:23:26Z
format Conference or Workshop Item
id nottingham-52186
institution University of Nottingham Malaysia Campus
institution_category Local University
last_indexed 2025-11-14T20:23:26Z
publishDate 2015
recordtype eprints
repository_type Digital Repository
spelling nottingham-521862020-05-04T17:15:20Z https://eprints.nottingham.ac.uk/52186/ AMP: a new time-frequency feature extraction method for intermittent time-series data Barrack, Duncan S. Goulding, James Hopcraft, Keith Preston, Simon Smith, Gavin The characterisation of time-series data via their most salient features is extremely important in a range of machine learning task, not least of all with regards to classification and clustering. While there exist many feature extraction techniques suitable for non-intermittent time-series data, these approaches are not always appropriate for intermittent time-series data, where intermittency is characterized by constant values for large periods of time punctuated by sharp and transient increases or decreases in value. Motivated by this, we present aggregation, mode decomposition and projection (AMP) a feature extraction technique particularly suited to intermittent time-series data which contain time-frequency patterns. For our method all individual time-series within a set are combined to form a non-intermittent aggregate. This is decomposed into a set of components which represent the intrinsic time-frequency signals within the data set. Individual time-series can then be _t to these components to obtain a set of numerical features that represent their intrinsic time-frequency patterns. To demonstrate the effectiveness of AMP, we evaluate against the real word task of clustering intermittent time-series data. Using synthetically generated data we show that a clustering approach which uses the features derived from AMP significantly outperforms traditional clustering methods. Our technique is further exemplified on a real world data set where AMP can be used to discover groupings of individuals which correspond to real world sub-populations. 2015-08-10 Conference or Workshop Item PeerReviewed Barrack, Duncan S., Goulding, James, Hopcraft, Keith, Preston, Simon and Smith, Gavin (2015) AMP: a new time-frequency feature extraction method for intermittent time-series data. In: 1st International Workshop on Mining and Learning from Time Series (MiLeTS), 10-13 August 2015, Sydney, Australia. time-series feature extraction intermittence
spellingShingle time-series
feature extraction
intermittence
Barrack, Duncan S.
Goulding, James
Hopcraft, Keith
Preston, Simon
Smith, Gavin
AMP: a new time-frequency feature extraction method for intermittent time-series data
title AMP: a new time-frequency feature extraction method for intermittent time-series data
title_full AMP: a new time-frequency feature extraction method for intermittent time-series data
title_fullStr AMP: a new time-frequency feature extraction method for intermittent time-series data
title_full_unstemmed AMP: a new time-frequency feature extraction method for intermittent time-series data
title_short AMP: a new time-frequency feature extraction method for intermittent time-series data
title_sort amp: a new time-frequency feature extraction method for intermittent time-series data
topic time-series
feature extraction
intermittence
url https://eprints.nottingham.ac.uk/52186/