Delta Divergence: A Novel Decision Cognizant Measure of Classifier Incongruence
- Submitting institution
-
The University of Surrey
- Unit of assessment
- 12 - Engineering
- Output identifier
- 9000586_4
- Type
- D - Journal article
- DOI
-
10.1109/TCYB.2018.2825353
- Title of journal
- IEEE Transactions on Cybernetics
- Article number
- -
- First page
- 2331
- Volume
- 49
- Issue
- 6
- ISSN
- 2168-2267
- Open access status
- Compliant
- Month of publication
- -
- Year of publication
- 2018
- URL
-
-
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
-
- Research group(s)
-
-
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Motivated by the deficiencies of Kullback-Leibler divergence as a measure for comparing the output of two classifiers, ( Ponti et.al.,Pattern Recognition 2017), Delta divergence, based on a completely novel concept - decision cognisant divergence - offers a new paradigm in measuring incongruence. Its superior theoretical properties have been complemented by experimental validation of its superior behaviour in the presence of estimation errors in Kittler et.al. (Pattern Recognition 2018). Its practical impact has been demonstrated on diverse applications (multimodal biometrics, video analytics). It also played a decisive role in securing a multimillion pound MURI project in Semantic Information Pursuit 2018-2023.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -