Evaluating the Validation of a Monte Carlo Simulation of Binary Time Series.
NAVAL RESEARCH LAB WASHINGTON DC
Pagination or Media Count:
A statistical technique has been developed for the validation of a Monte Carlo simulation process or other processes whose results can be reduced to a finite time sequence of equally spaced events with dichotomous outcomes. Essential to the technique is the Bahadur-Lazarsfeld representation of the probability distribution of the populations consisting of all binary vectors with a specific number of elements. This paper analyzes the properties of the test and the adequacy of the Bahadur-Lazarsfeld representation for practical application purposes. It examines the probability of rejection which results when the significance level alpha is specified and computes the power of the test when the null hypothesis is tested against known alternatives. The evidence collected suggests that the statistical test is an adequate tool for the purposes for which it was designed. Author
- Statistics and Probability
- Computer Programming and Software