On Finite Approximations to Markov Decision Processes with Recursive and Nonlinear Discounting
2021
In this paper, finite approximation schemes are justified for Markov decision processes in Borel spaces with recursive and nonlinear discounting. Explicit error bounds are obtained in terms of the system primitives. This allows one to solve the original problem approximately up to any given accuracy, by solving a sequence of problems in finite spaces.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
17
References
0
Citations
NaN
KQI