Feature Selection for Value Function Approximation Using Bayesian Model Selection.
Tobias JungPeter StonePublished in: ECML/PKDD (1) (2009)
Keyphrases
- bayesian model selection
- feature selection
- model selection
- posterior probability
- em algorithm
- data association
- parameter estimation
- state space
- feature set
- feature extraction
- machine learning
- unsupervised learning
- text classification
- expectation maximization
- mixture model
- feature space
- maximum likelihood
- probability distribution
- statistical analysis
- support vector
- training data
- three dimensional