The exact approximation method in distribution simulation.

Abstract

An algorithm for generating random variates quickly, called the exact approximation method is the subject of this thesis. Two other algorithms, the acceptance rejection method, and the inversion method for generating random variates are also included. The exact approximation method uses the acceptance rejection method. The inversion method is a special case of the exact approximation method. To generate random variates quickly using the exact approximation method a function which is a close approximation to the inverse cumulative probability function must be found. The choice of this function is a compromise between it's ease of computation and it's closeness to the inverse cumulative distribution function. Since both factors affect the efficiency of the exact approximation algorithm.