Change point detection in time series data with random forests
A large class of monitoring problems can be cast as the detection of a change in the parameters of astatic or dynamic system, based on the effects of these changes on one or more observed variables. Inthis paper, the use of random forest models to detect change points in dynamic systems is considere...
| Main Authors: | , |
|---|---|
| Format: | Journal Article |
| Published: |
Pergamon
2010
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/20.500.11937/28287 |
| _version_ | 1848752495557345280 |
|---|---|
| author | Auret, L. Aldrich, Chris |
| author_facet | Auret, L. Aldrich, Chris |
| author_sort | Auret, L. |
| building | Curtin Institutional Repository |
| collection | Online Access |
| description | A large class of monitoring problems can be cast as the detection of a change in the parameters of astatic or dynamic system, based on the effects of these changes on one or more observed variables. Inthis paper, the use of random forest models to detect change points in dynamic systems is considered.The approach is based on the embedding of multivariate time series data associated with normalprocess conditions, followed by the extraction of features from the resulting lagged trajectory matrix.The features are extracted by recasting the data into a binary classification problem, which can besolved with a random forest model. A proximity matrix can be calculated from the model and from thismatrix features can be extracted that represent the trajectory of the system in phase space. The resultsof the study suggest that the random forest approach may afford distinct advantages over a previouslyproposed linear equivalent, particularly when complex nonlinear systems need to be monitored. |
| first_indexed | 2025-11-14T08:09:32Z |
| format | Journal Article |
| id | curtin-20.500.11937-28287 |
| institution | Curtin University Malaysia |
| institution_category | Local University |
| last_indexed | 2025-11-14T08:09:32Z |
| publishDate | 2010 |
| publisher | Pergamon |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | curtin-20.500.11937-282872017-09-13T15:19:08Z Change point detection in time series data with random forests Auret, L. Aldrich, Chris - Singular value decomposition - Detection algorithms - Machine learning Time series analysis - Subspace methods A large class of monitoring problems can be cast as the detection of a change in the parameters of astatic or dynamic system, based on the effects of these changes on one or more observed variables. Inthis paper, the use of random forest models to detect change points in dynamic systems is considered.The approach is based on the embedding of multivariate time series data associated with normalprocess conditions, followed by the extraction of features from the resulting lagged trajectory matrix.The features are extracted by recasting the data into a binary classification problem, which can besolved with a random forest model. A proximity matrix can be calculated from the model and from thismatrix features can be extracted that represent the trajectory of the system in phase space. The resultsof the study suggest that the random forest approach may afford distinct advantages over a previouslyproposed linear equivalent, particularly when complex nonlinear systems need to be monitored. 2010 Journal Article http://hdl.handle.net/20.500.11937/28287 10.1016/j.conengprac.2010.04.005 Pergamon restricted |
| spellingShingle | - Singular value decomposition - Detection algorithms - Machine learning Time series analysis - Subspace methods Auret, L. Aldrich, Chris Change point detection in time series data with random forests |
| title | Change point detection in time series data with random forests |
| title_full | Change point detection in time series data with random forests |
| title_fullStr | Change point detection in time series data with random forests |
| title_full_unstemmed | Change point detection in time series data with random forests |
| title_short | Change point detection in time series data with random forests |
| title_sort | change point detection in time series data with random forests |
| topic | - Singular value decomposition - Detection algorithms - Machine learning Time series analysis - Subspace methods |
| url | http://hdl.handle.net/20.500.11937/28287 |