Learning Characteristics of Reverse Quaternion Neural Network

TL;DR

This paper introduces a novel multi-layer feedforward quaternion neural network architecture called Reverse Quaternion Neural Network, exploring its unique learning characteristics, including comparable learning speed and distinct rotation representation capabilities.

Contribution

It proposes the first study of reverse quaternion neural networks, highlighting their learning behavior and ability to represent rotations differently from existing models.

Findings

Comparable learning speed to existing quaternion networks

Ability to obtain different rotation representations

Potential for enhanced rotation modeling

Abstract

The purpose of this paper is to propose a new multi-layer feedforward quaternion neural network model architecture, Reverse Quaternion Neural Network which utilizes the non-commutative nature of quaternion products, and to clarify its learning characteristics. While quaternion neural networks have been used in various fields, there has been no research report on the characteristics of multi-layer feedforward quaternion neural networks where weights are applied in the reverse direction. This paper investigates the learning characteristics of the Reverse Quaternion Neural Network from two perspectives: the learning speed and the generalization on rotation. As a result, it is found that the Reverse Quaternion Neural Network has a learning speed comparable to existing models and can obtain a different rotation representation from the existing models.