Improving Offline Value-Function Approximations for POMDPs by Reducing Discount Factors.
Yi-Chun ChenMykel J. KochenderferMatthijs T. J. SpaanPublished in: IROS (2018)
Keyphrases
- linear approximation
- factors affecting
- real time
- factors that influence
- reinforcement learning
- learning algorithm
- efficient computation
- closed form
- multi agent
- dynamical systems
- markov decision processes
- domain independent
- finite state
- multi agent systems
- factors influencing
- key factors
- knowledge base
- information systems