Unit Stratified Sampling as a Tool for Approximation of Stochastic Optimization Problems
<p><p><span style="font-size: 11.000000pt; font-family: 'CMR10';">We apply stratified sampling with equiprobable strata and a single observa- tion drawn from each stratum to the approximate computation of stochastic programming problems. We determine the convergence rate of the approximation error both when com- puting expectations and when approximating stochastic programming problems.</span>