On the optimization of approximate control variates with parametrically defined estimators

Geoffrey F. Bomarito, Patrick E. Leser, James E. Warner, William P. Leser

Abstract: Highlights•Multi-model Monte Carlo methods, such as approximate control variates, can be used for estimator variance reduction.•The crux of the approximate control variates framework is sample allocation optimization.•Parametrically-defined approximate control variates are developed to improve sample allocation optimization.•Three newly defined estimators are developed and shown to improve variance reduction.