Nuclear Science and Engineering / Volume 48 / Number 2 / June 1972 / Page 219
Technical Note / dx.doi.org/10.13182/NSE72-A22473
Articles are hosted by Taylor and Francis Online.
Let α be a parameter governing a Monte Carlo importance sampling process. This note gives formulas by which the results of a Monte Carlo run may be used to estimate the derivatives of the variance with respect to α. It is pointed out that these derivatives may be used to optimize α, that is, to minimize the variance of the Monte Carlo problem.