A Normalized Probabilistic Expectation-Maximization Neural Network for Minimizing Bayesian Misclassification Cost Risk.
Parag C. PendharkarPublished in: Neural Process. Lett. (2013)
Keyphrases
- expectation maximization
- neural network
- misclassification costs
- cost sensitive
- probabilistic model
- generative model
- bayesian networks
- em algorithm
- maximum likelihood
- cost sensitive learning
- cost sensitive classification
- naive bayes
- mixture model
- training examples
- class distribution
- multi class
- image segmentation
- k means
- total cost
- data sets
- unsupervised learning
- similarity measure
- bayesian inference
- bayesian framework
- probability distribution
- binary classifiers
- active learning
- text mining
- training set
- graph cuts