Abstract
Recent research in genetically programmed polynomial neural networks demonstrates that they are successful in time-series modelling. Currently, an important issue is how to quantify the reliability of the inferred polynomial networks, that is how to evaluate their uncertainty in order to obtain evidence for their usefulness. This paper elaborates on approaches to estimation of error bars (confidence and prediction intervals) for polynomial neural networks including: 1) the analytical delta method, implemented using a neural network technique; 2) the empirical bootstrap method; and 3) an empirical network training method. We present results on empirical data which show that the delta method may lead to more unstable intervals and thus favour the bootstrap for practical applications.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Akaike, H.: Power Spectrum Estimation through Autoregression Model Fitting. Annals Inst. Stat. Math. 21, 407–419 (1969)
Box, G.E.P. and Jenkins, G.M.: Time Series Analysis Forecasting and Control. Holden-Day, San Francisco (1970)
Breiman, L.: Heuristics of Instability and Stabilization in Model Selection. The Annals of Statistics 24, N:6, 2350–2383 (1996)
Cawley, G.C. and Janacek, G.J.: Predictive Uncertainty in Environmental Modelling. Neural Networks 20, N:4, 537–549 (2007)
Chryssolouris, G., Lee, M. and Ramsey, A.: Confidence Interval Prediction for Neural Network Models. IEEE Trans. on Neural Networks 7, N:1, 229–232 (1996)
De Veaux, R.D., Schumi, J., Schweinsberg, J. and Lyle, H.U.: Prediction Intervals for Neural Networks via Nonlinear Regression. Technometrics 40, N:4, 273–282 (1998)
Dybovski, R. and Roberts, S.J.: Confidence Intervals and Prediction Intervals for Feedforward Neural Network Models, In: Dybovski,R. and Gant,V. (eds.) Clinical Applications of Artificial Neural Networks. pp.298-326. Cambidge University Press, UK, (2001)
Efron, B. and Tibshirani, R.J.: An Introduction to the Bootstrap. Chapman and Hall, New York, NY (1989)
Hwang, J.T.G. and Ding, A.A.: Prediction intervals for artificial neural networks. Journal of the American Statistical Association 92, N:438, 748–757 (1997)
Iba, H., Paul, T.K. and Hasegawa, Y.: Applied Genetic Programming and Machine Learning. CRC Press, Boca Raton, FL (2010)
Khosravi, A., Nahavandi, S., Creighton, D. and Atiya, A.F.: A Comprehensive Review of Neural Network-based Prediction Intervals and New Advances. IEEE Trans. on Neural Networks, (2011)
Koza, J.R.: Genetic Programming: On the Programming of Computers by Means of Natural Selection. The MIT Press, Cambridge, MA (1992)
Lauretto, M.S., Nakano, F., Pereira, C.A.B. and Stern, J.M.: Hierarchical Forecasting with Polynomial Nets, In: Nakamatsu,K. et al. (eds.) Proc. New Advances in Intelligent Decision Technologies, pp.305–315, Springer, Berlin (2009)
Nikolaev, N.Y. and Iba, H.: Adaptive Learning of Polynomial Networks: Genetic Programming, Backpropagation and Bayesian Learning. Springer, New York (2006)
Nix, D.A. and Weigend, A.S.: Estimating the Mean and Variance of the Target Probability Distribution. In: Proc. of the IEEE Int. Conf. on Neural Networks (IEEE-ICNN’94). pp.55–60. IEEE Press (1994)
Papadopoulos, G., Edwards, P.J. and Murray, A.F.: Confidence Estimation Methods for Neural Networks: A Practical Comparison. IEEE Trans. on Neural Networks 12, N:6, 1278–1287 (2001)
Pearlmutter, B.A.: Fast Exact Multiplication by the Hessian. Neural Computation 6, N:2, 147–160 (1994)
Rivals, I. and Personnaz, L.: Construction of Confidence Intervals for Neural Networks Based on Least Squares Estimation. Neural Networks 13, 463–484 (2000)
Rumelhart, D.E., Hinton, G.E., and Williams, R.J.: Learning Internal Representations by Error Propagation, In: Rumelhart, D., and McClelland, J. (eds.) Parallel Distributed Processing: Explorations in the Microstructure of Cognition. vol.1, 318–362 The MIT Press, Cambridge, MA (1986)
Seber, G.A.F. and Wild, C.: Nonlinear Regression. Wiley, New York, NY (1989)
Tibshirani, R.: A Comparison of Some Error Estimates for Neural Network Models. Neural Computation 8, N:1, 152–163 (1996)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2012 Springer Science+Business Media, LLC
About this paper
Cite this paper
Nikolaev, N., Smirnov, E. (2012). Error Bars for Polynomial Neural Networks. In: Dai, H., Liu, J., Smirnov, E. (eds) Reliable Knowledge Discovery. Springer, Boston, MA. https://doi.org/10.1007/978-1-4614-1903-7_3
Download citation
DOI: https://doi.org/10.1007/978-1-4614-1903-7_3
Published:
Publisher Name: Springer, Boston, MA
Print ISBN: 978-1-4614-1902-0
Online ISBN: 978-1-4614-1903-7
eBook Packages: Computer ScienceComputer Science (R0)