Title :
Spherical Neural Networks and some applications
Author :
Vallejo, Jose Refugio
Author_Institution :
Univ. of Guanajuato, Guanajuato
Abstract :
This paper presents a definition of spherical neural networks a natural extension to traditional neural networks using the framework of the geometric Clifford algebras. This extension allows us to describe appropriately the objects with spherical or cylindrical structure in the conformal geometric algebra. We present an application to separate information in the interior from information in the exterior of spherical or cylindrical structures.
Keywords :
algebra; geometry; neural nets; conformal geometric algebra; geometric Clifford algebras; spherical neural networks; Algebra; Computational efficiency; Computer networks; Equations; Matrices; Multidimensional systems; Neural networks; Particle separators; Physics; Solid modeling;
Conference_Titel :
Neural Networks, 2006. IJCNN '06. International Joint Conference on
Conference_Location :
Vancouver, BC
Print_ISBN :
0-7803-9490-9
DOI :
10.1109/IJCNN.2006.247011
