Abstract.
In this paper, inspired from our previous algorithm, which was based on the theory of Tsallis statistical mechanics, we develop a new evolving stochastic learning algorithm for neural networks. The new algorithm combines deterministic and stochastic search steps by employing a different adaptive stepsize for each network weight, and applies a form of noise that is characterized by the nonextensive entropic index q, regulated by a weight decay term. The behavior of the learning algorithm can be made more stochastic or deterministic depending on the trade off between the temperature T and the q values. This is achieved by introducing a formula that defines a time-dependent relationship between these two important learning parameters. Our experimental study verifies that there are indeed improvements in the convergence speed of this new evolving stochastic learning algorithm, which makes learning faster than using the original Hybrid Learning Scheme (HLS). In addition, experiments are conducted to explore the influence of the entropic index q and temperature T on the convergence speed and stability of the proposed method.
Similar content being viewed by others
References
S. Haykin, Neural Networks: A Comprehensive Foundation (Macmillan College Publishing Company, 1994)
G. Gyorgyi, Physics Reports 342, issue 4–5, 263 (2001)
S. Kirkpatrick, C.D. Gelatt Jr., M.P. Vecchi, Science 220, 671 (1983)
D. Ackley, G. Hinton, T. Sejnowski, Cogn. Sci. 9, 147 (1985)
E.H.L. Arts, J. Korst, Simulated Annealing and Boltzmann Machines (Wiley, New York, 1989)
R.M. Burton, G.J. Mpitsos, Neural Networks 5, 627 (1992)
T. Rögnvaldsson, Neural Computation 6, 916 (1994)
N.K. Treadgold, T.D. Gedeon, IEEE Tr. Neural Networks 9, 4, 662 (1998)
A.D. Anastasiadis, G.D. Magoulas, Physica A 344, 372 (2004)
M. Riedmiller, H. Braun, A direct adaptive method for faster backpropagation learning: The Rprop algorithm, Proc. Int. Conf. Neur. Net. (San Francisco, CA, 1993), pp. 586–591
C. Tsallis, D.A. Stariolo, Physica A 233, 395 (1996)
C. Tsallis, J. Stat. Phys. 52, 479 (1988)
H. Szu, Nonconvex optimization by fast simulated annealing, Proceedings of IEEE, Vol. 75 (1987), pp. 1538–1540
G. Snedecor, W. Cochran, Statistical Methods, 8th edn. (Iowa State University Press, 1989)
L. Prechelt, PROBEN1–A set of benchmarks and benchmarking rules for neural network training algorithms, Technical report 21/94, Fakultät für Informatik, Universität Karlsruhe, 1994
P.M. Murphy, D.W. Aha, UCI Repository of machine learning databases, http://www.ics.uci.edu/ mlearn/MLRepository.html, 1994
M.V. Boland, R.F. Murphy, IEEE Engineering in Medicine and Biology, Sept./Oct. (1999), pp. 115–119
H. Lodish, A. Berk, S.L. Zipursky, P. Matsudaira, D. Baltimore, J. James Darnell, Molecular Cell Biology, 5th edn. (Freeman, 2003)
P. Horton, K. Nakai, Better Prediction of Protein Cellular Localization Sites with the k Nearest Neighbors Classifier, Proc. of Intelligent Systems in Molecular Biology (1997), pp. 368–383
A.D. Anastasiadis, G.D. Magoulas, X. Liu, Classification of protein localisation patterns via supervised neural network learning, Proc. of the Fifth Symposium on Intelligent Data Analysis, Lecture Notes in Computer Science 2810, (Springer-Verlag, 2003), pp. 430–439
R. Kohavi, A study of cross-validation and bootstrap for accuracy estimation and model selection, International Joint Conference on Artificial Intelligence (1995), pp. 223–228
E.K. Blum, Neural Computation 1, 532 (1989)
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Anastasiadis, A., Magoulas, G. Evolving Stochastic Learning Algorithm based on Tsallis entropic index. Eur. Phys. J. B 50, 277–283 (2006). https://doi.org/10.1140/epjb/e2006-00137-6
Received:
Published:
Issue Date:
DOI: https://doi.org/10.1140/epjb/e2006-00137-6