N-rule genetic self-organizing map using genetic algorithm

Proposes a method for n-rule representation and n-rule inference. A network based on this method takes an architecture of a genetic self organizing map, called an n-rule genetic self-organizing map (n-RGSOM). The paper also provides n-nodes maintenance rules for n-RGSOM and utilizes some criteria for node overlapping prevention generated by the genetic operation. By using these criteria we can avoid critical area errors, and a node can take a single operation per iteration. The simulation results show that the genetic learning algorithm and n-nodes maintenance method developed in the paper is effective.