Advertisement

On ASMOD — An Algorithm for Empirical Modelling using Spline Functions

  • Tom Kavli
  • Erik Weyer
Chapter
Part of the Advances in Industrial Control book series (AIC)

Abstract

Empirical modelling algorithms build mathematical models of systems based on observed data. This chapter describes the theoretical foundation and principles of the ASMOD algorithm, including some improvements on the original algorithm. The ASMOD algorithm uses B-splines for representing general nonlinear models of several variables. The internal structure of the model is, through an incremental refinement procedure, automatically adapted to the dependencies observed in the data. Only input variables which are found of relevance are included in the model, and the dependency of different variables are decoupled when possible. This makes the model more parsimonious and also more transparent to the user. Case studies are included which confirm the usefulness of the algorithm.

Keywords

Gluconic Acid Multivariate Adaptive Regression Spline Minimum Description Length Structural Risk Minimisation Spline Space 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    H. Akaike. Fitting autoregressive models for prediction. Ann. Inst Stat. Math, 21:425–439, 1969.Google Scholar
  2. 2.
    H. Akaike. A new look at the statistical model identification. IEEE Trans. Automatic Control, 19:716–723, 1974.MathSciNetzbMATHCrossRefGoogle Scholar
  3. 3.
    T. Bohlin. Derivation of a designer’s guide for interactive ‘gray-box’ identification of nonlinear stochastic objects. Int. Journal of Control, 59(6):1505–1524, 1994.MathSciNetzbMATHCrossRefGoogle Scholar
  4. 4.
    N.A. Bridgett, M. Brown, C.J. Harris, and D.J. Mills. High-dimensional approximation using an associative memory network. In Proc. of IEE Control’94, volume 2, pages 1458 – 1465, 1994.Google Scholar
  5. 5.
    M. Brown, D.J. Mills, and C.J. Harris. The representation of fuzzy algorithms used in adaptive modelling and control schemes. Submitted to “Intelligent Systems Engineering”, 1994.Google Scholar
  6. 6.
    M. Brown and C. Harris. Neurofuzzy adaptive modelling and control. Prentice-Hall International (UK) Limited, 1994.Google Scholar
  7. 7.
    M. Carlin, T. Kavli, and B. Lillekjendlie. A comparison of four methods for nonlinear data modelling. Chemometrics and Intelligent Laboratory Systems, 23:163–177, 1994.CrossRefGoogle Scholar
  8. 8.
    R. Franke. Recent advances in the approximation of surfaces from scattered data. In C. K. Chui and L. L. Schumaker, editors, Topics in Multivariate Approximation, pages 79 – 98. Academic Press, Inc., New York, 1987.Google Scholar
  9. 9.
    J. Friedman. Multivariate adaptive regression splines. The Annals of Statistics, 19(1):1 – 141, 1991.MathSciNetzbMATHCrossRefGoogle Scholar
  10. 10.
    S. Geman, E. Bienenstock, and R. Doursat. Neural networks and the bias/variance dilemma. Neural Computation, 4:1–58, 1992.CrossRefGoogle Scholar
  11. 11.
    T.K. Ghose and P. Ghosh. Kinetic analysis of gluconic acid production by Pseudomonas ovalis. J. Applied Chemical Biotechnology, 26:768 – 777, 1976.CrossRefGoogle Scholar
  12. 12.
    T.J. Hastie and R.J. Tibshirani. Generalized Additive Models. Chapman and Hall, London, 1990.zbMATHGoogle Scholar
  13. 13.
    P.J. Huber. Robust Statistics. John Wiley and Sons, Inc., 1981.zbMATHCrossRefGoogle Scholar
  14. 14.
    D.R Hush and B.G. Home. Progress in supervised neural networks. what is new since Lippman. IEEE Signal Processing Magazine, pages 8 – 39, Jan 1993.Google Scholar
  15. 15.
    T.A. Johansen. Identification of non-linear systems using empirical data and prior knowledge — An optimization approach. Submitted to Automatica, Nov 1994.Google Scholar
  16. 16.
    T.A. Johansen and B.A. Foss. State-space modelling using operator regime decomposition and local models. Technical report 93–40-w, The Norwegian Institute of Technology, Trondheim, Norway, Jul 1993.Google Scholar
  17. 17.
    T.A. Johansen and B.A. Foss. Semi-empirical modelling of non-linear dynamic systems through identification of operating regimes and local models. In G. Irwin, K. Hunt and K. Warwick, editors, Advances in Neural Networks for Control Systems, Advances in Industrial Control. Springer-Verlag, 1995. This volume.Google Scholar
  18. 18.
    T. Kavli. ASMOD — an algorithm for adaptive spline modelling of observation data. International Journal of Control, 58(4):947–967, 1993.MathSciNetzbMATHCrossRefGoogle Scholar
  19. 19.
    T. Kavli. Learning principles in dynamic control. PhD thesis, University of Oslo, ISBN no. 82–411–0394–8, SINTEF, Oslo, 1993.Google Scholar
  20. 20.
    P. Lancaster and K. Šalkauskas. Curve and Surface Fitting. An introduction. Academic Press, Inc., London, 1986.zbMATHGoogle Scholar
  21. 21.
    I.J. Leontaritis and S.A. Billings. Model selection and validation methods for non-linear systems. Int. Journal of Control, 45(1):311–341, 1987.zbMATHCrossRefGoogle Scholar
  22. 22.
    J. Moody and C.J. Darken. Fast learning in networks of locally-tuned processing units. Neural Computation, 1:281 – 294, 1989.CrossRefGoogle Scholar
  23. 23.
    J. Park and I.W. Sandberg. Universal approximation using radial-basis-function networks. Neural Computation, 3(2):246–257, 1991.CrossRefGoogle Scholar
  24. 24.
    T. Poggio and F. Girosi. Networks for approximation and learning. Proceedings of the IEEE, 78(9):1481–1497, Sep 1990.CrossRefGoogle Scholar
  25. 25.
    M.J.D. Powell. Radial basis functions for mulivariable interpolation: A review. In J. C. Mason and M. G. Cox, editors, Algorithms for Approximation. Clarendon Press, London, 1987.Google Scholar
  26. 26.
    W.H. Press, B.P. Flannery, S.A. Teukolsky, and W.T. Vetterling. Numerical Recipes in C. Cambridge University Press, Cambridge, 1988.zbMATHGoogle Scholar
  27. 27.
    J. Rissanen. Modelling by shortest data description. Automatica, 14:465–471, 1978.zbMATHCrossRefGoogle Scholar
  28. 28.
    T.D. Sanger. A tree-structured algorithm for reducing computation in networks with separable basis functions. Neural Computation, 3(1):67–78, 1991.CrossRefGoogle Scholar
  29. 29.
    L.L. Schumaker. Spline Functions: Basic Theory. Wiley, New York, 1981.zbMATHGoogle Scholar
  30. 30.
    V. Vapnik. Estimation of Dependencies Based on Empirical Data. Springer-Verlag, 1982.Google Scholar
  31. 31.
    G. Wahba. Spline Models for Observational Data. Society for Industrial and Applied Mathematics, Philadelphia, Pennsylvania, 1990.zbMATHCrossRefGoogle Scholar
  32. 32.
    E. Weyer. System Identification in the Behavioral Framework. PhD thesis, The Norwegian Institute of Technology, Trondheim, Norway, 1992.Google Scholar
  33. 33.
    E. Weyer and T. Kavli. Theoretical and practical aspects of the asmod algorithm. Internal reports (in preparation), 1994.Google Scholar
  34. 34.
    E. Weyer, R.C. Williamson, and I.M.Y. Mareels. System identification in the behavioral framework via risk minimisation. In preparation, 1994.Google Scholar

Copyright information

© Springer-Verlag London Limited 1995

Authors and Affiliations

  • Tom Kavli
    • 1
  • Erik Weyer
    • 2
  1. 1.SINTEFOsloNorway
  2. 2.Department of Chemical EngineeringThe University of QueenslandBrisbaneAustralia

Personalised recommendations