An extreme function theory for novelty detection
We introduce an extreme function theory as a novel method by which probabilistic novelty detection may be performed with functions, where the functions are represented by time-series of (potentially multivariate) discrete observations. We set the method within the framework of Gaussian processes (GP...
| Main Authors: | , , , , |
|---|---|
| Format: | Journal article |
| Published: |
IEEE
2012
|
| _version_ | 1797074186646061056 |
|---|---|
| author | Clifton, D Clifton, L Hugueny, S Wong, D Tarassenko, L |
| author_facet | Clifton, D Clifton, L Hugueny, S Wong, D Tarassenko, L |
| author_sort | Clifton, D |
| collection | OXFORD |
| description | We introduce an extreme function theory as a novel method by which probabilistic novelty detection may be performed with functions, where the functions are represented by time-series of (potentially multivariate) discrete observations. We set the method within the framework of Gaussian processes (GP), which offers a convenient means of constructing a distribution over functions. Whereas conventional novelty detection methods aim to identify individually extreme data points, with respect to a model of normality constructed using examples of “normal” data points, the proposed method aims to identify extreme functions, with respect to a model of normality constructed using examples of “normal” functions, where those functions are represented by time-series of observations. The method is illustrated using synthetic data, physiological data acquired from a large clinical trial, and a benchmark time-series dataset. |
| first_indexed | 2024-03-06T23:32:41Z |
| format | Journal article |
| id | oxford-uuid:6c9ac841-e0e3-44fc-8ebb-324a363075fc |
| institution | University of Oxford |
| last_indexed | 2024-03-06T23:32:41Z |
| publishDate | 2012 |
| publisher | IEEE |
| record_format | dspace |
| spelling | oxford-uuid:6c9ac841-e0e3-44fc-8ebb-324a363075fc2022-03-26T19:11:59ZAn extreme function theory for novelty detectionJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:6c9ac841-e0e3-44fc-8ebb-324a363075fcSymplectic Elements at OxfordIEEE2012Clifton, DClifton, LHugueny, SWong, DTarassenko, LWe introduce an extreme function theory as a novel method by which probabilistic novelty detection may be performed with functions, where the functions are represented by time-series of (potentially multivariate) discrete observations. We set the method within the framework of Gaussian processes (GP), which offers a convenient means of constructing a distribution over functions. Whereas conventional novelty detection methods aim to identify individually extreme data points, with respect to a model of normality constructed using examples of “normal” data points, the proposed method aims to identify extreme functions, with respect to a model of normality constructed using examples of “normal” functions, where those functions are represented by time-series of observations. The method is illustrated using synthetic data, physiological data acquired from a large clinical trial, and a benchmark time-series dataset. |
| spellingShingle | Clifton, D Clifton, L Hugueny, S Wong, D Tarassenko, L An extreme function theory for novelty detection |
| title | An extreme function theory for novelty detection |
| title_full | An extreme function theory for novelty detection |
| title_fullStr | An extreme function theory for novelty detection |
| title_full_unstemmed | An extreme function theory for novelty detection |
| title_short | An extreme function theory for novelty detection |
| title_sort | extreme function theory for novelty detection |
| work_keys_str_mv | AT cliftond anextremefunctiontheoryfornoveltydetection AT cliftonl anextremefunctiontheoryfornoveltydetection AT huguenys anextremefunctiontheoryfornoveltydetection AT wongd anextremefunctiontheoryfornoveltydetection AT tarassenkol anextremefunctiontheoryfornoveltydetection AT cliftond extremefunctiontheoryfornoveltydetection AT cliftonl extremefunctiontheoryfornoveltydetection AT huguenys extremefunctiontheoryfornoveltydetection AT wongd extremefunctiontheoryfornoveltydetection AT tarassenkol extremefunctiontheoryfornoveltydetection |