1
An evaluation simulation of estimation methods for the Pearson type 3 distribution | |
Author | Chien, Chi-fu |
Call Number | AIT Thesis no. CS-87-3 |
Subject(s) | Distribution (Probability theory) |
Note | A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering, School of Engineering and Technology |
Publisher | Asian Institute of Technology |
Abstract | The Method of Mixed moments and other methods for parameter estimation in Pearson Type 3 distributions have been investigated in this study. A computer simulation program was conducted to evaluate the performance of twelve methods of different sample sizes. The major comparisons were for Method of Maximum Likelihood, Method of Direct moments and Method of Mixed moments. Censored samples were also examined in this study. Performance evaluation of the better methods had been made by using bias of parameter estimates, Root Mean Square Error(RMSE) of parameter estimates, and relative error and variance of T-year event corresponding to three probabilities levels. The comparison was tried to base on the same environment and the repeated count was large enough to allow a general conclusion. It showed that the Method of Direct Moment has good performance in tenms of relative error of T-year event over small sample size while the Method of Maximum Likelihood has good performance over large sample sizes. But in general, MLTl of Method of Mixed moments performs well for both small and large samples. |
Year | 1987 |
Type | Thesis |
School | School of Engineering and Technology (SET) |
Department | Other Field of Studies (No Department) |
Academic Program/FoS | Computer Science (CS) |
Chairperson(s) | Huynh, Ngoc Phien |
Examination Committee(s) | Khang, Bach Hung ;Venkatachalam, Paruvachi A. |
Scholarship Donor(s) | Telecommunication Administration Taiwan, Republic of China |
Degree | Thesis (M.Eng.) - Asian Institute of Technology, 1987 |