Simulating Discounted Costs.
Abstract
In many settings, discounted costs arise naturally. This paper describes simulation methodologies for estimation of expected discounted costs associated with systems that exhibit stochastic fluctuations. Such techniques are important for numerical computation of discounted costs for stochastic processes in which conventional numerical methods either fail to apply or are inefficient. Examples of such processes include non-Markov processes or infinite state space Markov chains. The discussion given here of simulation algorithms for the discounted cost problem also merits interest to the extent that it provides an excellent vehicle for illustrating several sophisticated variance reduction methods for stochastic simulation. These techniques are more accurately called efficiency increase techniques. Keywords: Truncation algorithm; Monte Carlo estimators.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1987
- Accession Number
- ADA193584
Entities
People
- Bennett L. Fox
- Peter W. Glynn
Organizations
- Stanford University