Abstract
The use of Genetic Algorithms (GA's) in Economic Simulation Models has been introduced by Holland (Holland, 1988) and justified by pointing out the similarities between Economics, Game Theory, Control Theory and Evolutionary Genetics Footnote 1 Unfortunately, such a justification is not sufficient in the long run, as similarities between different terms do not necessarily justify the transformation of methods and mechanisms from one domain to the other. The justification problem might be regarded as aggravated by the fact that some researchers - among others (Paul, 1993); (Boehme, 1993) - have been using Genetic Algorithms in a straightforward manner within Economic Simulations whereas Holland's original use of GA's is restricted to rule-generation within Classifier Systems. What therefore is required is the introduction of a justification based on economic theory as a base for such simulations.
* Corresponding author.
1 “It is traditional in economics to carry out this characterization by attaching a utility to the various states of the environment. The role of utility in economics is quite similar to the role of payoffin game theory, the error function in control theory, fitness in evolutionary genetics, and so on.” (Holland, 1988, p. 120).
* Corresponding author.
1 “It is traditional in economics to carry out this characterization by attaching a utility to the various states of the environment. The role of utility in economics is quite similar to the role of payoffin game theory, the error function in control theory, fitness in evolutionary genetics, and so on.” (Holland, 1988, p. 120).
Notes
* Corresponding author.
1 “It is traditional in economics to carry out this characterization by attaching a utility to the various states of the environment. The role of utility in economics is quite similar to the role of payoffin game theory, the error function in control theory, fitness in evolutionary genetics, and so on.” (Holland, 1988, p. 120).