Comparison of Update and Genetic Training Algorithms in a Memristor Crossbar Perceptron

TL;DR

This paper compares local update and genetic training algorithms for memristor crossbar neural networks, demonstrating differences in resilience to hardware failures during image classification tasks.

Contribution

It introduces a comparison of two training algorithms' robustness to memristor hardware failures in neural network implementations.

Findings

Genetic algorithm shows higher resilience to memristor failures.

Local update scheme's training ability degrades faster with increasing failures.

Genetic training maintains higher accuracy under hardware failure conditions.

Abstract

Memristor-based computer architectures are becoming more attractive as a possible choice of hardware for the implementation of neural networks. However, at present, memristor technologies are susceptible to a variety of failure modes, a serious concern in any application where regular access to the hardware may not be expected or even possible. In this study, we investigate whether certain training algorithms may be more resilient to particular hardware failure modes, and therefore more suitable for use in those applications. We implement two training algorithms -- a local update scheme and a genetic algorithm -- in a simulated memristor crossbar, and compare their ability to train for a simple image classification task as an increasing number of memristors fail to adjust their conductance. We demonstrate that there is a clear distinction between the two algorithms in several measures of the rate of failure to train.