Maximum Likelihood Estimation
in the Three-parameter Log-normal Distribution Using the Continuation
Method
Authors
H. Hirose
Source
Computational Statistics
& Data Analysis, Vol.24, No.2, pp.139-152 (1997.4)
Abstract
The main purpose of this paper is to show an extremely
successful maximum likelihood parameter estimation scheme for the
three-parameter log-normal distribution. The proposed algorithm,
which is a combination of the continuation method and the extended
log-normal distribution model, can find the existing local maximum
likelihood estimates successfully without a careful selection of
the starting point in the iterative process.