Methods for the General Analysis of Genetic Algorithms as Economic Learning Techniques

Economic models examining social learning with the help of GAs always come in the form of simulations. For a precisely specified economic problem, a computer program is constructed. The most important part of this program is the formulation of artificial economic agents, which are often constructed as objects. The most important part of the agents, in turn, is their decision system which, in this work, consists of a genetic algorithm or at least of parts of a GA. The core analysis of the behavior of the agents and by that the behavior of the economic model consists of running the simulation program, recording the resulting data and analyzing and interpreting the data. All the results won from this work are inductive by nature. Even if the simulation is run very many times: The results do not allow for deductive conclusions, they are never true in general, but only in a set of special circumstances.