DESPOT: Online POMDP Planning with Regularization.
Adhiraj SomaniNan YeDavid HsuWee Sun LeePublished in: NIPS (2013)
Keyphrases
- planning problems
- online learning
- partially observable markov decision processes
- belief space
- real time
- belief state
- partially observable
- partially observable stochastic domains
- stochastic domains
- state space
- dynamical systems
- heuristic search
- planning process
- multi agent
- finite state
- blocks world
- planning under uncertainty
- reinforcement learning
- motion planning
- domain independent
- optimal planning
- partial observability
- continuous state
- predictive state representations