Control variate (CV) estimators are a useful class of reduced variance estimators for simulation experiments. They correct the value of a crude estimator after the simulation outputs have been realized, and thus they do not alter the time path of the simulated stochastic process. In this paper, a broad characterization of CV estimators is proposed, the properties of this class of estimators are investigated, and the characterization is employed to derive new CV estimators that have advantageous properties in addition to reduced variance.
ASJC Scopus subject areas
- Computer Science(all)
- Modeling and Simulation
- Management Science and Operations Research