Online discriminative kernel density estimator with Gaussian kernels
- Submitting institution
-
The University of Birmingham
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 24112805
- Type
- D - Journal article
- DOI
-
10.1109/TCYB.2013.2255983
- Title of journal
- IEEE Transactions on Cybernetics
- Article number
- -
- First page
- 355
- Volume
- 44
- Issue
- 3
- ISSN
- 2168-2267
- Open access status
- Out of scope for open access requirements
- Month of publication
- March
- Year of publication
- 2014
- 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
-
1
- Research group(s)
-
-
- Citation count
- 22
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- The paper proposes a new method for supervised online estimation of probabilistic discriminative models for classification tasks. The main novelty lies in a new cost function that measures loss of interclass discrimination during compression and guides the compression toward simpler models that still retain discriminative properties. This is significant as it has been demonstrated that we can achieve comparable classification performance of best batch methods while allowing online adaptation with models of lower complexity. The method has been applied by many researchers, most notably by the group of Prof Tuytelaars (KU Leuven), and is referenced in US Patent https://patents.justia.com/patent/10446368.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -