Sample Average Approximation Method for Chance Constrained ...

distribution in chance constraint by an empirical distribution corresponding to a randomsample. This approach is well known for stochastic programs with expected valuesobjectives [9]. SAA methods for chance constrained problems have been investigatedin [10] and [11].The remainder of the paper is organized as follows. In Section 2 we provide theoreticalbackground for the SAA approach, showing convergence of the optimal value of theapproximation to the optimal value of the true problem. In addition, following [8] wedescribe how to construct bounds for the optimal value of chance constrained problemsof the form (1). In Section 3, we present a chance constrained portfolio selection problem.We apply the SAA method to obtain upper bounds as well as candidate solutionsto the problems. In addition we present several numerical experiments that indicatehow one should tune the parameters of the SAA approach. In Section 4 we presenta simple blending problem modeled as a joint chance constrained problem. Section 5concludes the paper and suggests directions for future research.We use the following notation throughout the paper. The integer part of numbera ∈ R is denoted by ⌊a⌋. By Φ(z) we denote the cumulative distribution function (cdf)of standard normal random variable and by z α the corresponding critical value, i.e.,4