On the generative-discriminative tradeoff approach: Interpretation, asymptotic efficiency and classification performance.
Jing-Hao XueD. M. TitteringtonPublished in: Comput. Stat. Data Anal. (2010)
Keyphrases
- discriminative learning
- bayesian network classifiers
- unsupervised learning
- feature extraction
- discriminative training
- conditional likelihood
- classification accuracy
- discriminative methods
- discriminative classifiers
- pattern recognition
- pattern classification
- image classification
- feature selection
- support vector
- machine learning
- generative model
- semi supervised
- class labels
- automatic classification
- discriminative models
- category labels
- discriminative information
- data sets
- generative and discriminative models
- supervised learning
- support vector machine
- computational complexity
- probabilistic interpretation
- neural network
- discriminative random fields
- probabilistic boosting tree
- decision trees
- discriminatively trained
- face recognition
- training set
- structured output
- dimension reduction
- text classification
- maximum likelihood
- em algorithm
- class specific
- classification method
- classification models