AN OPTIMALITY CONDITION FOR DISCRETE DYNAMIC PROGRAMMING WITH NO DISCOUNTING,
Abstract
The report presents a discussion of undiscounted problems having infinite planning horizons. The study develops an optimality condition for the discrete time, finite stage Markov decision problem with Veinott's criterion of maximizing the Cesaro mean of the vector returns in a finite horizon as the horizon tends to infinity. Veinott's conjecture that there are optimal stationary policies is also verified.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1967
- Accession Number
- AD0663406
Entities
People
- B. L. Miller
- E. V. Denardo
Organizations
- RAND Corporation