Abstract
The purpose of this study is to decide an appropriate neighborhood and a transition rule of cellular automata by analyzing the past growth process of urban areas in Busan. The fuzzy set conception is applied to minimize the loss of data, and the Analytic Hierarchy Process(AHP) is used to estimate the relative weight of each factor. With cellular automata which can manage the change based on the dynamic model and time, this study analyzes the urban growth of Busan from 1987 to 1997. Also, through the simulation of different types for neighborhood and transition rules, we can find the appropriate neighborhood and the transition rule for Busan. It is the sensitivity analysis, Cramer's V and Kappa calculation to verify the accuracy of the simulation. In conclusion, the forecast of physical urban growth pattern is more accurate under conditions when the number of matrixes for the neighborhood is small, the shape of the neighborhood is circular, "" value, which control the pace of urban growth, is high and the transition possibility () is low.