In this thesis, we investigated the performance of Cultural Algorithms over the complete range of system complexities, from fixed to chaotic. Specifically we were interested in observing the following: (I) Whether there was a similar process going on in the solution of a problem, regardless of the complexity of the problem, we developed a simulation environment, the Cones World, within which to express representative complexity class examples as suggested by Langton. (II) Whether there was a given homogeneous topology that could dominate across all complexities, (III) Whether we can monitor the "vital signs" of a cultural system during the search process to determine whether it was on track or not, (IV) And finally can we infer the complexity class of an organization based on its vital signs.In order to apply the Cultural Algorithm over all complexity classes it was necessary that we generalize on its co-evolutionary nature in order to keep the variation in the population across all complexities. As a result we produced a new version of the Cultural Algorithms Toolkit, CAT 2.0, which supported a variety of co-evolutionary features at both the Knowledge and Population levels.We then applied the system to the solution of a 150 randomly generation problems. As a result, we were able to produce the following conclusion: (I) No homogeneous Social fabric tested was dominant over all categories of complexity. (II) As the complexity of problems increased so did the complexity of the Social Fabric that was need to deal with it efficiently. So a fabric that was good for fixed problems would be less adequate for periodic problems, and chaotic ones. (III) There was a basic thermodynamic metaphor that described how the knowledge sources interacted in a successful search. This metaphor was called the Cultural Engine.The metaphor was described in terms of measures used to describe the entropy of each of the Cultural Algorithm components. The influence function worked as Maxwell's Demon to inject new entropy into the system so as to counteract the effect of the second law of thermodynamics. We then used this model to interpret the successful runs given for the system. The results suggest that the Cultural Engine has the potential to be a powerful metaphor for problem solving in social systems.
展开▼
