Compact kernel classifiers trained with minimum classification error criterion.
Ryoma TaniHideyuki WatanabeShigeru KatagiriMiho OhsakiPublished in: MLSP (2017)
Keyphrases
- minimum classification error
- kernel classifiers
- discriminative training
- statistical pattern recognition
- hidden markov models
- maximum likelihood
- probabilistic interpretation
- maximum likelihood estimation
- decision functions
- classification error
- support vector machine
- posterior probability
- training set
- gaussian mixture
- perceptron algorithm
- active learning
- pairwise
- decision rules
- svm classifier
- support vector
- training data
- co training
- machine learning
- feature vectors
- face recognition