Effect of memristor\'s potentiation-depression curves peculiarities in the convergence of physical perceptrons

TL;DR

This study investigates how the peculiarities of memristor potentiation-depression curves influence the convergence speed of physical perceptrons, with implications for optimizing memristor-based neural networks.

Contribution

It provides experimental P-D curves for manganite memristors and analyzes their impact on perceptron learning convergence, including effects of non-linearity, asymmetry, and noise.

Findings

Lower granularity P-D curves reduce convergence time.

Non-linear and asymmetric P-D responses improve learning speed.

Noise injection further decreases convergence epochs.

Abstract

Neuromorphic computing aims to emulate the architecture and information processing mechanisms of the mammalian brain. This includes the implementation by hardware of neural networks. Oxide-based memristor arrays with cross-bar architecture appear as a possible physical implementation of neural networks.In this paper, we obtain experimental potentiation-depression (P-D) curves on different manganite-based memristive systems and simulate the learning process of perceptrons for character recognition. We analyze how the specific characteristics of the P-D curves affect the convergence time -- characterized by the EPOCHs-to-convergence (ETC) parameter -- of the network. Our work shows that ETC is reduced for systems displaying P-D curves with relatively low granularity and non-linear and asymmetric response. In addition, we also show that noise injection during the synaptic weight actualization further reduces the ETC. The results obtained here are expected to contribute to the optimization of hardware neural networks based on memristors cross-bar arrays.