An Introduction to Quaternion-Valued Recurrent Projection Neural Networks

TL;DR

This paper introduces quaternion-valued recurrent projection neural networks (QRPNNs), which enhance associative memory capabilities by overcoming cross-talk issues and improving storage capacity and noise tolerance compared to previous quaternion neural networks.

Contribution

The paper presents QRPNNs, a novel combination of non-local projection learning with quaternion-valued recurrent neural networks, addressing key limitations of prior models.

Findings

QRPNNs overcome the cross-talk problem of QRCNNs.

QRPNNs exhibit greater storage capacity than QRCNNs.

QRPNNs demonstrate improved noise tolerance.

Abstract

Hypercomplex-valued neural networks, including quaternion-valued neural networks, can treat multi-dimensional data as a single entity. In this paper, we introduce the quaternion-valued recurrent projection neural networks (QRPNNs). Briefly, QRPNNs are obtained by combining the non-local projection learning with the quaternion-valued recurrent correlation neural network (QRCNNs). We show that QRPNNs overcome the cross-talk problem of QRCNNs. Thus, they are appropriate to implement associative memories. Furthermore, computational experiments reveal that QRPNNs exhibit greater storage capacity and noise tolerance than their corresponding QRCNNs.