Non-Stationary Dynamic Programming with Additive and Multiplicative Rewards,
Abstract
The author considers a non-stationary, discrete time, stochastic dynamic programming model in which the problem is to determine a policy for choosing actions that maximizes the expected total reward. The novelty of this model is that it covers multiplicative rewards as well as the usual additive rewards. The optimality equation, a value iteration procedure and related results are studied for Borel state and action spaces and essentially negative or positive rewards. Sufficient conditions for the existence of deterministic optimal policies are presented for uniformly bounded rewards.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1975
- Accession Number
- ADA013138
Entities
People
- Robert C. Wang
Organizations
- Syracuse University