PJO: Evaluation of scenario generation methods for stochastic programming



		Volume 3, Number 2, May 2007, pp. 257-271
		Michal Kaut and Stein W. Wallace
	Key words:
	stochastic programming, scenario tree, scenario generation, stability
Mathematices Subject Classification: 90C15, 0C31
	References


	Abstract:
	Stochastic programs can only be solved with discrete distributions of limited cardinality. Input, however, normally comes in the form of continuous distributions or large data sets. Creating a limited discrete distribution from input is called scenario generation. In this paper, we discuss how to evaluate the quality or suitability of scenario generation methods for a given stochastic programming model. We formulate minimal requirements that should be imposed on a scenario generation method before it can be used for solving the stochastic programming model. We also show how the requirements can be tested. The procedures for testing a scenario generation method is illustrated on a case from portfolio management.

Evaluation of scenario generation methods for stochastic programming