Abstract
In this paper, we study the incorporation of Bayesian Regularization into constructive neural networks. The degree of regularization is automatically controlled in the Bayesian inference framework and hence does not require manual setting. Simulation shows that regularization with input training using a full Bayesian approach produces networks with better recognition performance and lower susceptibility as the noise increases. Regularization with input training under the gradient descent algorithm, however, does not produce significant improvement on the problems tested.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
MacKay, D.J.C.: Bayesian Interpolation. Neural Computation (1991)
Skilling, J.: Maximum Entropy and Bayesian Methods, pp. 45–52. Kluwer Academic Publishers, Dordrecht (1989)
Wang, Z.W., ZhangLiu, R.P.: One Improved BP Neural Network Learning Algorithm. Mathematical Theory and Applications 25, 31–34 (2005)
Abu-Mostafa, Y.S.: Learning from Hints in Neural Networks. J. Complexity 6, 192–198 (1990)
MacKay, D.J.C.: Bayesian Interpolation. Neural Computation 4(3), 415–447 (1992)
Tishby, N., Levin, E., Solla, S.A.: Consistent Inference of Probabilities in Layered Networks: Predictions and Generalization. In: Proc. IJCNN, Washington (1989)
MacKay, D.J.C.: A Practical Bayesian Framework for Back Propagation Networks. Neural Computation 4(3), 448–472 (1992)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2009 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Huang, X., Zeng, H. (2009). English Letters Recognition Based on Bayesian Regularization Neural Network. In: Wang, H., Shen, Y., Huang, T., Zeng, Z. (eds) The Sixth International Symposium on Neural Networks (ISNN 2009). Advances in Intelligent and Soft Computing, vol 56. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-01216-7_37
Download citation
DOI: https://doi.org/10.1007/978-3-642-01216-7_37
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-01215-0
Online ISBN: 978-3-642-01216-7
eBook Packages: EngineeringEngineering (R0)