Realizing Hebbian Learning Rule on a Hardware

Abstract

The approaches developed for intelligent systems are not only inspired from cognitive processes, there are also structures based on the working principles of the brain. Lately, these structures comprise a large amount of studies related to artificial intelligence. The structures developed are largely realized in a software environment. This gives rise to high cost apparently in time and energy. Thus, it is important to generate special purpose hardware. In this work, based on the mathematical model of neuron, Hebbian rule, which is proposed by Hebb for the interaction between two neurons and is related to learning at cell level, is realized on hardware. So, a case for the hardware realization of a learning rule, which is used in constituting autonomous learning systems in artificial intelligence studies, is given.