A theory for learning based on rigid bodies dynamics

Author

Fiori, Simone

Author_Institution

Neural Networks & Adaptive Syst. Res. Group, Perugia Univ., Italy

Volume

13

Issue

3

fYear

2002

fDate

5/1/2002 12:00:00 AM

Firstpage

521

Lastpage

531

Abstract

A new learning theory derived from the study of the dynamics of an abstract system of masses, moving in a multidimensional space under an external force field, is presented. The set of equations describing system´s dynamics may be directly interpreted as a learning algorithm for neural layers. Relevant properties of the proposed learning theory are discussed within the paper, along with results of computer simulations performed in order to assess its effectiveness in applied fields

Keywords

computational complexity; learning (artificial intelligence); principal component analysis; signal processing; computer simulations; external force field; independent component analysis; learning theory; neural layers; orthonormal signal processing; principal component analysis; rigid bodies dynamics; unsupervised neural learning; Biomedical signal processing; Computer simulation; Direction of arrival estimation; Independent component analysis; Multidimensional signal processing; Multidimensional systems; Neural networks; Nonlinear dynamical systems; Nonlinear equations; Signal processing algorithms;

fLanguage

English

Journal_Title

Neural Networks, IEEE Transactions on

Publisher

ieee

ISSN

1045-9227

Type

jour

DOI

10.1109/TNN.2002.1000121

Filename

1000121