Members
Overall Objectives
Research Program
Application Domains
Software and Platforms
New Results
Bilateral Contracts and Grants with Industry
Partnerships and Cooperations
Dissemination
Bibliography
XML PDF e-pub
PDF e-Pub


Section: Software and Platforms

Dissimilarity Coefficient learning

Participant : Pawan Kumar [Correspondant] .

weakly supervised learning, dissimilarity coefficient, structured prediction Disc (publicly available at http://cvn.ecp.fr/personnel/pawan/code/DISCAPI.zip ) software provides a convenient API for dissimilarity coefficient (DISC) based learning. DISC allows the use of weakly supervised datasets (with missing information) by jointly learning a structured prediction classifier and a conditional probability distribution of the missing information. The parameters of the classifier and the distribution are learned by minimizing a user-specified dissimilarity coefficient between them.