earticle

영어

This paper discusses the use of new graph structural genetic programming for automatic programming, which creates finite state machines (FSM) by evolution. Generally, FSM must
define their transition rules for all combinations of states and possible inputs, thus the FSM
program will become large and complex when the number of states and inputs is large. In our
work, the nodes are connected by trajectory information sets, so it is possible that only the
essential problem’s behavior obtained in the current situation are used in the network flow,
and it can determine an action by not only the current, but also the past information. In addition, the proposed algorithm enhances evolutionary process by using fitness inheritance technique. Constraining the depth of genetic programming tree is one of the ways to overcome its bloat problem. Finally, fitness inherent is used when fitness evaluation is computationally expensive. Fitness inherent is based on averaging; therefore it reflects some assumptions of smoothness in the search space