No Access

A NEW STRUCTURE AND TRAINING PROCEDURE FOR MULTI-MOTHER WAVELET NETWORKS

Faculty of Sciences, University of Gafsa, 2112 Gafsa, Tunisia

REGIM: REsearch Group on Intelligent Machines, National School of Engineers (ENIS), University of Sfax, 3018 Sfax, Tunisia

Search for more papers by this author

, and

ADEL M. ALIMI

REGIM: REsearch Group on Intelligent Machines, National School of Engineers (ENIS), University of Sfax, 3018 Sfax, Tunisia

Search for more papers by this author

https://doi.org/10.1142/S0219691310003353Cited by:24 (Source: Crossref)

Abstract

This paper deals with the features of a new wavelet network structure founded on several mother wavelets families. This new structure is similar to the classic wavelets network but it admits some differences eventually. The wavelet network basically uses the dilations and translations versions of only one mother wavelet to construct the network, but the new one uses several mother wavelets and the objective is to maximize the probability of selection of the best wavelets. Two methods are presented to assist the training procedure of this new structure. On one hand, we have an optimal selection technique that is based on an improved version of the Orthogonal Least Squares method; on the other, the Generalized Cross-Validation method to determine the number of wavelets to be selected for every mother wavelet. Some simulation results are reported to demonstrate the performance and the effectiveness of the new structure and the training procedure for function approximation in one and two dimensions.

Keywords:

References

A. M. Alimi, The beta-based neuro-fuzzy system: Genesis and main properties, TASK Quarterly Journal, Special Issue on "Neural Networks" eds. W. Duch and D. Rutkowska 7(1) (2003) 23–41 . Google Scholar
B. Thomas and C. E. Joseph, Comput. Stat. Data Anal. 50(10), 2635 (2006). Web of Science, Google Scholar
B. Yevqeniyet al., Expert Systems 22(5), 235 (2005). Web of Science, Google Scholar
C. Ben Amar, W. Bellil and A. M. Alimi, Trans. Syst. Signals Devices 1(3), 275 (2006). Web of Science, Google Scholar
C. E. Vasioset al., Cross-validation and neural network architecture selection for the classification of intracranial current sources, Proc. 2004 Seventh Seminar on Neural Network Applications in Electrical Engineering, NEUREL 2004 (2004) pp. 151–158. Google Scholar
C.-J. Lin, Applied Soft Computing 9, 71 (2009), DOI: 10.1016/j.asoc.2008.03.014. Crossref, Web of Science, Google Scholar
C. Kai and H. Yoshio, An adaptive hybrid wavelet neural network and its application, Proc. 2004 IEEE International Conference on Robotics and Biomimetics, IEEE ROBIO 2004 (2004) pp. 73–86. Google Scholar
D. G. Luenberger , Linear and Nonlinear Programming ( Addison-Wesley , 1996 ) . Google Scholar
D. Whitley, T. Starkweather and C. Bogart, Parallel Computing 14, 347 (1990), DOI: 10.1016/0167-8191(90)90086-O. Crossref, Web of Science, Google Scholar
F. Gruau and D. Whitley, Evolutionary Computation 1(3), 213 (1993), DOI: 10.1162/evco.1993.1.3.213. Crossref, Web of Science, Google Scholar
F. Rossi and C. Gegout, Geometrical initialization, parametrization and control of multilayer perceptrons: Application to function approximation, Proc. WCCI ICNNI (1994) pp. 546–550. Google Scholar
G. H. Golub and C. F. Van Loan , Matrix Computations , 2nd edn. ( Johns Hopkins University Press , 1989 ) . Google Scholar
H. Akaike, IEEE Trans. Automat. Control AC 19(6), 716 (1974), DOI: 10.1109/TAC.1974.1100705. Crossref, Web of Science, Google Scholar
H. M. Abbas, J. Computers 4(2), 147 (2009). Web of Science, Google Scholar
H. Min and C. Baotong, A novel learning algorithm for wavelet neural networks, First International Conference on Natural Computation, ICNC (2005) pp. 1–17. Google Scholar
H. Q. Thuan and S. Rudy, Effective neural network pruning using cross validation, Proc. International Joint Conference on Neural Networks (2005) pp. 972–977. Google Scholar
H. R. Karimiet al., Int. J. Wavelets Multiresolut. Inf. Process. 4(1), 213 (2006), DOI: 10.1142/S0219691306001178. Link, Web of Science, Google Scholar
J. Lewalle, Int. J. Wavelets Multiresolut. Inf. Process. 7(1), 131 (2009), DOI: 10.1142/S0219691309002738. Link, Web of Science, Google Scholar
S.-J. Kim, J.-Y. Seo and H.-T. Jeon, Design of the scaling-wavelet neural network using genetic algorithm, Proc. Int. Joint Conference on Neural Networks (2002) pp. 2174–2179. Google Scholar
I. Daubechies , Ten Lectures on Wavelets ( Society for Industrial and Applied Mathematics , Philadelphia, Pennsylvania , 1992 ) . Crossref, Google Scholar
C. G. Lim, K. Kim and E. Kim, Modeling for an adaptive wavelet network parameter learning using genetic algorithms, Proc. Fifteenth IASTED International Conference on Modeling and Simulation (2004) pp. 55–59. Google Scholar
L. Wessels and E. Barnard, IEEE Trans. Neural Networks 3(6), 899 (1992), DOI: 10.1109/72.165592. Crossref, Web of Science, Google Scholar
Y. Liuet al., A new model selection criterion based on information geometry, 2004 Seventh International Conference on Signal Processing Proceedings, ICSP (2004) pp. 1561–1564. Google Scholar
M. Mehra and B. V. Rathish Kumar, Int. J. Wavelets Multiresolut. Inf. Process. 5(4), 667 (2007), DOI: 10.1142/S0219691307001963. Link, Web of Science, Google Scholar
M. K. Titsias and A. C. Likas, IEEE Trans. Neural Network 12(5), 987 (2001), DOI: 10.1109/72.950129. Crossref, Web of Science, Google Scholar
O. Yacine and D. Gerard, Neurocomputing 34, 131 (2002). Crossref, Web of Science, Google Scholar
Q. Zhang and A. Benveniste, IEEE Trans. Neural Networks 3(6), 889 (1992), DOI: 10.1109/72.165591. Crossref, Web of Science, Google Scholar
Q. Zhang, IEEE Trans. Neural Networks 8(2), 227 (1997). Crossref, Web of Science, Google Scholar
R. Bulirsch and J. Stoer , Introduction to Numerical Analysis ( Springer-Verlag , 1980 ) . Google Scholar
R. H. Franke, A critical comparison of some methods for interpolation of scattered data, Tech. Rep. NPS-53-79-003, Dept. of Mathematics, Naval Postgraduate School, Monterey, Calif., USA (1979) . Google Scholar
R. J. Renka, ACM Trans. Math. Software 14(2), 139 (1988), DOI: 10.1145/45054.45055. Crossref, Web of Science, Google Scholar
R. Yan and R. X. Gao, Int. J. Wavelets Multiresolut. Inf. Process. 7(4), 411 (2009), DOI: 10.1142/S0219691309002994. Link, Web of Science, Google Scholar
S. A. Billings and H. L. Wei, IEEE Trans. Neural Networks 16(4), 862 (2005), DOI: 10.1109/TNN.2005.849842. Crossref, Web of Science, Google Scholar
S. Haykin , Neural Networks: A Comprehensive Foundation ( Macmillan , 1994 ) . Google Scholar
Q. Sui and Y. Gao, A stepwise updating algorithm for multiresolution wavelet neural networks, Proc. International Conference on Wavelet Analysis and Its Applications (WAA) (2003) pp. 633–638. Google Scholar
V. Colla, L. M. Reyneri and M. Sgarbi, Orthogonal least squares algorithm applied to the initialization of multi-layer perceptrons, ESANN'1999 Proc. European Symposium on Artificial Neural Networks (1999) pp. 363–369. Google Scholar
W. Bellil, C. Ben Amar and A. M. Alimi, Beta wavelet based image compression, International Conference on Signal, System and Design, SSD031 (2003) pp. 77–82. Google Scholar
W. Bellil, M. Othmani and C. Ben Amar, Initialization by selection for multi-library wavelet neural network training, in Proc. 3rd International Workshop on Artificial Neural Networks and Intelligent Information Processing, ANNIIP07, in conjunction with 4th International Conference on Informatics in Control, Automation and Robotics, ICINCO, Anger, France (2007), pp. 30–37 . Google Scholar
Y. Akifumi, N. Mamoru and I. Makio, Comm. Statist. Theory Methods 34(12), 2331 (2005). Web of Science, Google Scholar
Y. Sheng, D. Roberge and H. H. Szu, Optical Engr. 31(9), 1840 (1992), DOI: 10.1117/12.59916. Crossref, Web of Science, Google Scholar
Y. S. Wang and C.-M. Lee, Int. J. Wavelets Multiresolut. Inf. Process. 7(4), 459 (2009), DOI: 10.1142/S0219691309003033. Link, Web of Science, Google Scholar
Z. Michalewicz, Evolutionary Computation 1(1), 51 (1993), DOI: 10.1162/evco.1993.1.1.51. Crossref, Web of Science, Google Scholar
T. Koji, S. Masashi and M. Klaus-Robert, IEEE Trans. Neural Networks 13(1), 70 (2002). Web of Science, Google Scholar
N. Murata, S. Yoshizawa and S. Amari, IEEE Trans. Neural Networks 5(6), 865 (1994), DOI: 10.1109/72.329683. Crossref, Web of Science, Google Scholar
N. T. Van Der Merwe and A. J. Hoffman, Developing an efficient cross validation strategy to determine classifier performance (CVCP), Proc. Int. Joint Vol. Conf. Neural Networks3 (2001) pp. 1663–1668. Google Scholar
J. Xu and D. W. C. Ho, A constructive algorithm for wavelet neural networks, First International Conference on Natural Computation (2005) pp. 730–739. Google Scholar
J. Zhang, G. G. Walter and Y. Miao, IEEE Trans. Signal Process. 43(6), 1485 (1995). Crossref, Web of Science, Google Scholar
Q. Zhang and A. Benvenist, IEEE Trans. Signal Process. 13(6), 889 (1992). Google Scholar
Q. Zhang, IEEE Trans. Neural Networks 8, 227 (1997). Crossref, Web of Science, Google Scholar