Approximate Planning in POMDPs with Weighted Graph Models.
Yong LinXingjia LuFillia MakedonPublished in: Int. J. Artif. Intell. Tools (2015)
Keyphrases
- point based value iteration
- partially observable markov decision processes
- partially observable
- belief space
- stochastic domains
- belief state
- reinforcement learning
- predictive state representations
- planning problems
- weighted sum
- approximate solutions
- partial observability
- markov decision processes
- neural network
- state space
- initial state
- planning process
- planning domains
- decision theoretic
- decision problems
- dynamical systems
- plan generation
- markov decision problems
- motion planning