Bio-inspired adaptive networks based on organic memristors

Abstract

The organic memristor is an element which varies its conductance according to its previous involvement in the signal transfer processes, i.e. it combines conductance with memory properties. The first part of the work is dedicated to the consideration of its basic principles and fundamental properties. After this, we present the architecture of the organization of model networks, demonstrating the capabilities of supervised and unsupervised learning. Finally, we discuss the possible ways, alternative to the existing lithography-based technologies, that would result in the fabrication of statistically organized networks of such elements, mimicking learning in biological systems.