For Full-Text PDF, please login, if you are a member of IEICE,|
or go to Pay Per View on menu list, if you are a nonmember of IEICE.
A New Approach to the Structural Learning of Neural Networks
Rameswar DEBNATH Haruhisa TAKAHASHI
IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences
Publication Date: 2004/06/01
Print ISSN: 0916-8508
Type of Manuscript: LETTER
Category: Neural Networks and Bioengineering
backpropagation, structural learning, Lagrangian differential gradient,
Full Text: PDF>>
Structural learning algorithms are obtained by adding a penalty criterion (usually comes from the network structure) to the conventional criterion of the sum of squared errors and applying the backpropagation (BP) algorithm. This problem can be viewed as a constrained minimization problem. In this paper, we apply the Lagrangian differential gradient method to the structural learning based on the backpropagation-like algorithm. Computational experiments for both artificial and real data show that the improvement of generalization performance and the network optimization are obtained applying the proposed method.