Share Email Print

Proceedings Paper

A heuristic neural network initialization scheme for modeling nonlinear functions in engineering mechanics: continuous development
Author(s): Jin-Song Pei; Eric C. Mai
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

This paper introduces a continuous effort towards the development of a heuristic initialization methodology for constructing multilayer feedforward neural networks to model nonlinear functions. In this and previous studies that this work is built upon, including the one presented at SPIE 2006, the authors do not presume to provide a universal method to approximate arbitrary functions, rather the focus is given to the development of a rational and unambiguous initialization procedure that applies to the approximation of nonlinear functions in the specific domain of engineering mechanics. The applications of this exploratory work can be numerous including those associated with potential correlation and interpretation of the inner workings of neural networks, such as damage detection. The goal of this study is fulfilled by utilizing the governing physics and mathematics of nonlinear functions and the strength of the sigmoidal basis function. A step-by-step graphical procedure utilizing a few neural network prototypes as "templates" to approximate commonly seen memoryless nonlinear functions of one or two variables is further developed in this study. Decomposition of complex nonlinear functions into a summation of some simpler nonlinear functions is utilized to exploit this prototype-based initialization methodology. Training examples are presented to demonstrate the rationality and effciency of the proposed methodology when compared with the popular Nguyen-Widrow initialization algorithm. Future work is also identfied.

Paper Details

Date Published: 10 April 2007
PDF: 12 pages
Proc. SPIE 6529, Sensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems 2007, 65291T (10 April 2007); doi: 10.1117/12.715956
Show Author Affiliations
Jin-Song Pei, Univ. of Oklahoma (United States)
Eric C. Mai, Univ. of Oklahoma (United States)

Published in SPIE Proceedings Vol. 6529:
Sensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems 2007
Masayoshi Tomizuka; Chung-Bang Yun; Victor Giurgiutiu, Editor(s)

© SPIE. Terms of Use
Back to Top