The simulating modeling by dynamical mechanism with perceptron and stochastic logic

This paper discusses how to apply a hybrid computation process comprised by three techniques of Artificial Neural Network (ANN), Bayesian probability and Cellular Automata (CA), to establish simulating model which can embody the advantages of perceptron, stochastic and dynamic. Through our hybrid computation process, we can perform a series of dynamic simulations with high accuracy. Additionally, we can evaluate the commercial values of street locations by the distribution of the transitional probabilities, and learn where streets are mutable with the impact of economic fluctuations by the analysis of the average of changing rates.