Extreme value statistics for novelty detection in biomedical signal processing
Extreme value theory (EVT) is a branch of statistics which concerns the distributions of data of unusually low or high value i.e. in the tails of some distribution. These extremal points are important in many applications as they represent the outlying regions of normal events against which we may w...
| Main Authors: | , |
|---|---|
| Format: | Conference item |
| Published: |
IEE
2000
|
| _version_ | 1797055658911072256 |
|---|---|
| author | Roberts, S IEE IEE |
| author_facet | Roberts, S IEE IEE |
| author_sort | Roberts, S |
| collection | OXFORD |
| description | Extreme value theory (EVT) is a branch of statistics which concerns the distributions of data of unusually low or high value i.e. in the tails of some distribution. These extremal points are important in many applications as they represent the outlying regions of normal events against which we may wish to define novel events. The use of such novelty detection approaches is useful for analysis of data for which few exemplars of some important class exist, for example in medical screening. It is shown that a principled approach to the issue of novelty detection may be taken using extreme value statistics. |
| first_indexed | 2024-03-06T19:12:56Z |
| format | Conference item |
| id | oxford-uuid:175e7299-18c6-487c-bf53-d4445c3dd7e1 |
| institution | University of Oxford |
| last_indexed | 2024-03-06T19:12:56Z |
| publishDate | 2000 |
| publisher | IEE |
| record_format | dspace |
| spelling | oxford-uuid:175e7299-18c6-487c-bf53-d4445c3dd7e12022-03-26T10:36:53ZExtreme value statistics for novelty detection in biomedical signal processingConference itemhttp://purl.org/coar/resource_type/c_5794uuid:175e7299-18c6-487c-bf53-d4445c3dd7e1Symplectic Elements at OxfordIEE2000Roberts, SIEEIEEExtreme value theory (EVT) is a branch of statistics which concerns the distributions of data of unusually low or high value i.e. in the tails of some distribution. These extremal points are important in many applications as they represent the outlying regions of normal events against which we may wish to define novel events. The use of such novelty detection approaches is useful for analysis of data for which few exemplars of some important class exist, for example in medical screening. It is shown that a principled approach to the issue of novelty detection may be taken using extreme value statistics. |
| spellingShingle | Roberts, S IEE IEE Extreme value statistics for novelty detection in biomedical signal processing |
| title | Extreme value statistics for novelty detection in biomedical signal processing |
| title_full | Extreme value statistics for novelty detection in biomedical signal processing |
| title_fullStr | Extreme value statistics for novelty detection in biomedical signal processing |
| title_full_unstemmed | Extreme value statistics for novelty detection in biomedical signal processing |
| title_short | Extreme value statistics for novelty detection in biomedical signal processing |
| title_sort | extreme value statistics for novelty detection in biomedical signal processing |
| work_keys_str_mv | AT robertss extremevaluestatisticsfornoveltydetectioninbiomedicalsignalprocessing AT iee extremevaluestatisticsfornoveltydetectioninbiomedicalsignalprocessing AT iee extremevaluestatisticsfornoveltydetectioninbiomedicalsignalprocessing |