01368nas a2200205   4500000000100000000000100001008004100002260000900043653002900052653002000081653002200101653003900123653001500162100001700177700001600194245006400210300001000274490000600284520087200290       2002                            d  c200210aBayesian Decision-Making10aDecision-making10aFractal Dimension10aintelligent agent simulation model10ameta-model1 aJames Moffat1 aSusan Witty00aPhase changes in Meta-modelling using the Fractal Dimension  a52-670 v83 a<p>We discuss in this paper the development of a meta-model of an intelligent agent simulation model. Such intelligent agent models consist of a number of entities called agents, which interact with each other. The nature of such interactions creates emergent behaviour. In conflict or peacekeeping situations, these agents correspond to the actors in the situation (the different force elements, for example). A meta-model is thus a mathematical abstraction of such a simulation, composed of two parts. For the first part, the fractal dimension of a force is introduced as a parameter measuring the emergent ability of such forces to cluster locally, corresponding to local decision-making by individual agents. For the second part we consider the mathematics of Bayesian Decision-Making as a meta-model for top down decision processes in such simulation models.</p>