The Problem of Experimental Design in Simulation
Abstract
In many cases, a simulation study may be viewed as an experimental situation in which a number of factors (independent variables) are to be investigated. However, standard experimental design techniques often require more simulation runs than are available to the simulation user. In general, the primary problem of experimental design in simulation can be succinctly summarized as too many factors and too few runs. A discussion of this problem is presented in this report. Three possible two-stage strategies for attacking the problem are considered, and performance measures with which to judge the strategies are described. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 1981
- Accession Number
- ADA099703
Entities
People
- Carl A. Mauro
- Dennis E. Smith