Generation of Correlated Pseudo-Random Numbers for Monte Carlo Simulation.

The purpose of this thesis is to develop specific analytical techniques which can be used to generate correlated sets of pseudo-random numbers from a specified joint normal distribution. A number of alternative transformation techniques are developed and discussed. The problem of generating correlated sets of zeroone Bernoulli pseudo-random numbers is also addressed. Techniques which are widely used for the generation of uncorrelated sets of pseudo-random numbers are briefly reviewed. These existing pseudo-random number generators are used to generate uncorrelated pseudo-random numbers which can be transformed to the desired set of correlated pseudo-random numbers. Author