IMPLANT: An Integrated MDP and POMDP Learning AgeNT for Adaptive Games.
Chek Tien TanHo-Lun ChengPublished in: AIIDE (2009)
Keyphrases
- learning agent
- reinforcement learning
- reward function
- learning agents
- state space
- markov decision processes
- stochastic games
- markov decision process
- optimal policy
- markov decision problems
- partially observable
- reinforcement learning algorithms
- learning capabilities
- average reward
- finite state
- partially observable markov decision processes
- solving problems
- function approximation
- nash equilibria
- generative model
- belief state
- transition probabilities
- multiple agents
- learning algorithm
- markov chain
- dynamic programming
- single agent
- heuristic search
- multi agent
- partially observable markov decision process
- planning problems
- decision problems
- dynamical systems
- utility function
- domain independent
- policy iteration
- transfer learning