Pairwise clustering based on the mutual-information criterion.
Amir AlushAvishay FriedmanJacob GoldbergerPublished in: Neurocomputing (2016)
Keyphrases
- information criterion
- pairwise
- model selection
- cross validation
- probability model
- bayesian information criterion
- factor analysis
- markov random field
- semi supervised
- higher order
- widely applicable
- sample size
- similarity measure
- belief propagation
- variable selection
- free energy
- face recognition
- discriminant analysis
- learning theory
- metric learning
- selection criterion
- statistical model
- pairwise constraints
- language model
- support vector machine
- leave one out cross validation