Model of symbolic looking procedure acquisition process in navigation learning task

In this paper we propose a model of symbolic procedure formation in the brain and its control architecture. In the model, a procedure in the brain is composed of multiple functional parts that realize subfunctions of the required computation. An internal attention vector sequence selects and activates these functional areas, and the processing circuit is realized. To evaluate the plausibility of the model, a computer simulation for acquisition of the reinforcement learning procedure for a navigation task is conducted. Two types of agents with different functional parts have acquired the learning procedure, and the relation of the result with the development of individual is discussed.