Average Case Analysis of the Classical Algorithm for Markov Decision Processes with Büchi Objectives.
Krishnendu ChatterjeeManas JoglekarNisarg ShahPublished in: FSTTCS (2012)
Keyphrases
- average case
- markov decision processes
- worst case
- dynamic programming
- learning algorithm
- model based reinforcement learning
- search space
- optimal policy
- policy iteration
- computational complexity
- state space
- optimal solution
- uniform distribution
- machine learning
- transition matrices
- factored mdps
- state abstraction
- average reward
- belief state
- finite state
- reinforcement learning
- sufficient conditions