Lognormal Distribution Maximum-Likelihood Parameter Estimation Algorithms

摘要

Microcomputer-based algorithms for lognormal distribution parameter estimation are presented. A program diskette will be made available upon request. For given data, the parameters (three or two) are estimated (1) by means of the moments, and (2) by means of the maximum-likelihood principle. Maximum-likelihood estimation is done in two different ways: (1) by means of the derivative equations which result from the logarithmic likelihood function, and (2) by means of direct optimization of the logarithmic likelihood function itself. Moment estimates are used as starting values for the optimization process. (Author).