A method of reducing the variance of an estimate obtained via simulation. Suppose, for example, that we wish to estimate the quantity I, given bywhere g is a given function. An obvious method is to generate pseudo-random numbers u1, u2,…, un, in the interval (0, 1). Writing U1, U2,…, Un for the corresponding random variables, the estimator, I1, is given byImportance sampling makes a more representative choice of values: suppose that f is a probability density function that resembles g in its general shape, and let F be the corresponding distribution function. Instead of working with U1, U2,…, Un, we work with V1, V2,…, Vn, where F(Vj)=Uj. The resulting estimator, I2, given bywill have a smaller variance than I1.