Robustness of classification ability of spiking neural networks

TL;DR

This paper investigates the robustness of spiking neural networks' classification ability under sinusoidal and Gaussian input perturbations, showing they maintain performance despite such disturbances.

Contribution

It provides the first extensive analysis of spiking neural network robustness against sinusoidal and Gaussian perturbations using the SpikeProp algorithm.

Findings

Classification ability remains stable under sinusoidal perturbations.

Gaussian perturbations do not significantly impair performance.

Experimental results on benchmark datasets support robustness claims.

Abstract

It is well-known that the robustness of artificial neural networks (ANNs) is important for their wide ranges of applications. In this paper, we focus on the robustness of the classification ability of a spiking neural network which receives perturbed inputs. Actually, the perturbation is allowed to be arbitrary styles. However, Gaussian perturbation and other regular ones have been rarely investigated. For classification problems, the closer to the desired point, the more perturbed points there are in the input space. In addition, the perturbation may be periodic. Based on these facts, we only consider sinusoidal and Gaussian perturbations in this paper. With the SpikeProp algorithm, we perform extensive experiments on the classical XOR problem and other three benchmark datasets. The numerical results show that there is not significant reduction in the classification ability of the network if the input signals are subject to sinusoidal and Gaussian perturbations.