Toward Breaking the Curse of Dimensionality: An FPTAS for Stochastic Dynamic Programs with Multidimensional Actions and Scalar States.
Nir HalmanGiacomo NanniciniPublished in: SIAM J. Optim. (2019)
Keyphrases
- stochastic dynamic
- initial state
- perceptual aliasing
- state transitions
- action sequences
- partial knowledge
- internal states
- reinforcement learning
- goal directed
- multi dimensional
- multidimensional data
- vector valued
- state information
- multidimensional space
- state transition
- pareto optimal
- reasoning about actions
- computer programs
- plan recognition
- learning algorithm