-meteorologists problem and its application to structure learning and feature selection in reinforcement learning.
Carlos DiukLihong LiBethany R. LefflerPublished in: ICML (2009)
Keyphrases
- structure learning
- reinforcement learning
- feature selection
- transfer learning
- markov blanket
- bayesian networks
- graphical models
- text categorization
- conditional independence
- multi task
- machine learning
- bayesian network classifiers
- markov logic networks
- parameter estimation
- bayesian network structure learning
- structure learning algorithm
- parameter learning
- state space
- feature space
- text classification
- probabilistic model
- network structure
- k nearest neighbor
- random variables
- model selection
- feature set
- learning process
- learning algorithm
- naive bayes
- maximum likelihood estimation
- sample size
- text mining
- support vector machine
- classification accuracy
- active learning