An automata network reverse engineering algorithm using system statics and dynamics

Abstract

An automata network reverse engineering algorithm that considers not only the change events (dynamics) but also their no-change events (statics) is proposed. The time complexity for the algorithm is theta(m(2)n(2)) where m is the number of given global system states and n is the number of system components. Experiments on stock exchange data showed that the extracted state transition rules for the components may reveal some hidden relations coming from statics of the system. The extracted automaton information about the interacting set of stocks is particularly valuable as it can be used for index tracking purposes.