IRMA-International.org: Creator of Knowledge
Information Resources Management Association
Advancing the Concepts & Practices of Information Resources Management in Modern Organizations

Artificial Neural Networks: Enhanced Back Propagation in Character Recognition

Artificial Neural Networks: Enhanced Back Propagation in Character Recognition
View Free PDF
Author(s): Evon M. Abu-Taieh (The Arab Academy for Banking and Financial Science, Jordan)
Copyright: 2003
Pages: 3
Source title: Information Technology & Organizations: Trends, Issues, Challenges & Solutions
Source Editor(s): Mehdi Khosrow-Pour, D.B.A. (Information Resources Management Association, USA)
DOI: 10.4018/978-1-59140-066-0.ch067
ISBN13: 9781616921248
EISBN13: 9781466665330

Abstract

The initial weights in Back Propagation has a strong influence in the learning speed and the quality of the solution obtained after convergence. The main objective of the research is to improve the performance of BP through three concepts: First, propagating the error back to the weights without squaring; second, limiting the initial randomly-generated weights to a certain range; and third, starting the matrix of weights with an ID matrix.

Body Bottom