The Use of the Empirical Probability Generating Function to Estimate the Neyman Type A Distribution Parameters.
NAVAL POSTGRADUATE SCHOOL MONTEREY CA
Pagination or Media Count:
The Maximum Likelihood estimators for the Neyman Type A distribution parameters are very difficult to compute. In this thesis, the Empirical Probability Generating Function is used to provide estimators that are easier to compute and have asymptotic efficiency at least as high as 97 of that for the Maximum Likelihood estimators over most of the parameter space considered. The estimators found by this method are consistently better than the Method of Moments and the Method of Zero Frequency estimators with respect to asymptotic efficiency. The considerations of preference in using one method over another are discussed. Author
- Statistics and Probability