Modeling of Spatial Navigation Inspired by Rodent Hippocampus

This paper solves spatial navigation task inspired by methods of living organisms, especially by hippocampus that is responsible for spatial navigation. We attempt to model the behavior of a rodent hippocampus by methods of artificial neural networks and reinforcement learning theory. A model of mutual relations among several parts of hippocampal formation is suggested. To investigate the behavior of our model a task similar to the well-known Morris water maze task is considered [8]. Key-Words: Hippocampus, spatial navigation, reinforcement learning, place cells, neural networks, animat.