Self-Organizing Visual Cortex Model using Homeostatic Plasticity Mechanism

Homeostatic plasticity is regulatory mechanism that modulates both synaptic connections and intrinsic properties of individual neuron in order to maintain stability of neuronal activity. This type of plasticity is considered to be an important complement to Hebbian plasticity, especially during learning and development process of nervous system. In this project, the role of homeostatic plasticity in self-organizing visual cortex model (LISSOM) will be investigated. Two computational model of homeostatic plasticity, gradient rule for intrinsic plasticity proposed by Triesch [27] and homeostatic synaptic scaling proposed by Sullivan and de Sa [26] will be applied to LISSOM model of V1. Several experiments will be conducted to evaluate the self-organization process and the organized map of the model with homeostatic mechanism. The result of this initial exploration indicates that homeostatic plasticity, particularly intrinsic plasticity, can be successfully integrated into LISSOM model of visual cortex to improve some of its behavior and make it more biologically realistic. ii Acknowledgement