A Study of Neural Network Based on Quaternion

Tomoaki Kusakabe; Teijiro Isokawa; Noriaki Kouda; Nobuyuki Matsui

doi:10.11499/sicep.2002.0.158.0

SICE Annual Conference 2002

DOI https://doi.org/10.11499/sicep.2002.0.158.0

Conference information

Host: The Society of Instrument and Control Engineers

Co-host: IEEE/Industiral Electronic Society, IEEE/Robotics and Automation Society, IEEE/Control System Society

A Study of Neural Network Based on Quaternion

Tomoaki Kusakabe, Teijiro Isokawa, Noriaki Kouda, Nobuyuki Matsui

Author information

Keywords: Quaternion, Quaternion BP network, Affine transformation

CONFERENCE PROCEEDINGS FREE ACCESS

Pages 158

Details

Abstract

A quaternion neural network with nonlinear function is presented in this paper. The quaternion neuron adopts geometrical transformations as operators of data to be processed. Back Propagation for the layered network with quaternion neurons is also formulated. The simulation results show that the presented model has the ability in learning affine transformation of 3D figures, while conventional neural network can hardly acquire.

Corresponding author

Register with J-STAGE for free!