Genetic algorithms for estimation problems with multiple opt

The genetic algorithm is examined as a method for solving optimization problems in econometric estimation. It does not restrict either the form or regularity of the objective function, allows a reasonably large parameter space, and does not rely on a point-to-point search. The performance is evaluat...

Full description

Saved in:
Bibliographic Details
Published inJournal of business & economic statistics Vol. 13; no. 1; p. 53
Main Authors Dorsey, Robert E, Mayer, Walter J
Format Journal Article
LanguageEnglish
Published Alexandria Taylor & Francis Ltd 01.01.1995
Subjects
Algorithms
Econometrics
Economic theory
Estimating techniques
Genetic algorithms
Optimization
Studies
Online AccessGet full text

Cover

More Information
Summary:The genetic algorithm is examined as a method for solving optimization problems in econometric estimation. It does not restrict either the form or regularity of the objective function, allows a reasonably large parameter space, and does not rely on a point-to-point search. The performance is evaluated through 2 sets of experiments on standard test problems as well as econometric problems from the literature. First, alternative genetic algorithms that vary over mutation and crossover rates, population rates, and other features are contrasted. Second, the genetic algorithm is compared to Nelder-Mead simplex, simulated annealing, adaptive random search, and MSCORE. Although the genetic algorithm has difficulty finding a point at which the gradient is exactly zero, it can greatly enhance the performance of locally efficient gradient-type algorithms by providing needed global efficiency. For nondifferentiable problems, the genetic algorithm can complement existing software packages such as MSCORE and can provide a flexible direct-search method for certain estimators.
ISSN:0735-0015
1537-2707