Abstract
Although Artificial Neural Networks (ANNs) have been extensively used to solve forecasting problems, defining their architectures has commonly been a very difficult task. Self-Organizing Polynomial Neural Networks can be used to alleviate this problem. However, it causes an increase in the computational cost and the addition of other parameters. This first drawback can be mitigated by using a matrix inversion technique as training algorithm, while the second, by using Differential Evolution. The method developed in this study combines those techniques in order to simultaneously search for the best parameters, the network architecture and weights. Finally, one can observe that in most databases the proposed method outperformed the Backpropagation, the most commonly used training algorithm in ANNs.
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
Ivakhnenko, A.G.: Polynomial theory of complex systems. IEEE Trans. on Systems, Man and Cybernetics 1, 364–378 (1971)
Oh, S.K., Pedrycz, W.: The design of self-organizing polynomial neural networks. Information Science 141, 237–258 (2002)
Oh, S.K., Pedrycz, W., Park, B.J.: Polynomial neural networks architecture: analysis and design. Computers and Electrical Engineering 29, 703–725 (2003)
Storn, R., Price, K.: Differential evolution–a simple efficient adaptive scheme for global optimization over continuous spaces. Technical Report 95-012, Int. Compt. Sci. Inst., Berkeley, CA (1995)
Zarth, A.M.F.: Otimização evolucionária multimodal de redes neurais artificiais com evolução diferencial. Master’s thesis of Federal University of Pernambuco (2010)
Frank, A., Asuncion, A.: Uci machine learning repository
Chambers, J.: The r project for statistical computing
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2012 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Tablada, L.G.N., Valença, M.J.S. (2012). Self-Organizing Polynomial Neural Networks Based on Matrix Inversion and Differential Evolution. In: Yin, H., Costa, J.A.F., Barreto, G. (eds) Intelligent Data Engineering and Automated Learning - IDEAL 2012. IDEAL 2012. Lecture Notes in Computer Science, vol 7435. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-32639-4_49
Download citation
DOI: https://doi.org/10.1007/978-3-642-32639-4_49
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-32638-7
Online ISBN: 978-3-642-32639-4
eBook Packages: Computer ScienceComputer Science (R0)