International Journal of Computer Mathematics, vol. 78, pp. 521-537, 2001 (SCI, IF 0.858)
A. S. Rodionov, H. Choo, H. Y. Youn, and T. M. Chung
In this paper a new algorithm for generating a random process with a given marginal distribution and autocorrelation function is presented. By using the truncated distribution instead of the order statistics used in the earlier algorithm, the proposed algorithm achieves significant speed up and accuracy. We also develop a scheme for deciding the transition probability matrix based on the Newton optimization technique, that is one of the key steps in the proposed algorithm. The experiment for 16 state randomized Markov chain shows about 14 times acceleration in comparison with the earlier algorithm. The storage requirement is also much smaller.
Autocorrelation; Newton optimization; Random process; Randomized Markov chain; Simulation