Generalised Entropy MDPs and Minimax Regret.
Emmanouil G. AndroulakisChristos DimitrakakisPublished in: CoRR (2014)
Keyphrases
- minimax regret
- reward function
- markov decision processes
- utility function
- preference elicitation
- decision problems
- reinforcement learning
- optimal policy
- state space
- stochastic programming
- misclassification costs
- transition probabilities
- data sets
- decision theory
- text classification
- linear programming
- cost sensitive
- class distribution
- dynamic programming
- training and test data
- decision making