Neighborhood Selection and Rules Identification for Cellular Automata: A Rough Sets Approach

摘要: In this paper a method is proposed which uses data mining techniques based on rough sets theory to select neighborhood and determine update rule for cellular automata (CA). According the approach, detected by reducts calculations rule-learning algorithm applied induce set of decision rules that define evolution CA. Experiments were performed with use synthetic as well real-world sets. The results show introduced allows identification both deterministic probabilistic CA-based models phenomena.