| EPSRC Reference: |
GR/K51334/01 |
| Title: |
NOVELTY DETECTION |
| Principal Investigator: |
Tarassenko, Professor L |
| Other Investigators: |
|
| Researcher Co-Investigators: |
|
| Project Partners: |
|
| Department: |
Engineering Science |
| Organisation: |
University of Oxford |
| Scheme: |
Standard Research (Pre-FEC) |
| Starts: |
01 October 1995 |
Ends: |
31 December 1997 |
Value (£): |
168,980
|
| EPSRC Research Topic Classifications: |
|
| EPSRC Industrial Sector Classifications: |
| Aerospace, Defence and Marine |
Healthcare |
|
| Related Grants: |
|
| Panel History: |
|
|
Summary on Grant Application Form |
|
There are many pattern recognition problems in engineering and medicine for which the class of real interest (the abnormalities or pathological cases) is under-represented in the database of available examples. As a result of this, this class is ignored when a neural network classifier is trained using minimisation of the mean-squared error at the output. Although it might be possible with some problems to compensate for this to a certain extent, abnormalities tend to be very rare in comparison to the examples from the normal classes and an altogether different approach is needed. The solution proposed here is to learn a description of novelty using neural network techniques and subsequently identify abnormalities by testing for novelty against the description of normality. The question which we seek to answer in the proposal is: how different is a previously unseen data vector from the vectors which form part of the training set of normal data?
|
|
Key Findings |
|
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
|
|
Potential use in non-academic contexts |
|
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
|
|
Impacts |
|
| Description |
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk |
| Summary |
|
| Date Materialised |
|
|
|
|
Sectors submitted by the Researcher |
|
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk
|
| Project URL: |
|
| Further Information: |
|
| Organisation Website: |
http://www.ox.ac.uk |