A Hilbert Space Approach to Variance Reduction

Elsevier Handbooks in Operations Research and Management Science: Simulation, pp 259-289.In this chapter we explain variance reduction techniques from the Hilbert space standpoint, in the terminating simulation context. We use projection ideas to explain how variance is reduced, and to link different variance reduction techniques. Our focus is on the methods of control variates, conditional Monte Carlo, weighted Monte Carlo, stratification, and Latin hypercube sampling