Advertisement

Speeding Up Back Propagation by Partial Evaluation

  • Henrik Friborg Jacobsen
  • Carsten Krogh Gomard
  • Peter Sestoft
Conference paper

Abstract

We automatically specialize a general Back Propagation learning algorithm to a particular network topology, obtaining a specialized learning algorithm which is faster than the general one.

The automatic specialization is done by a partial evaluator for a subset of the imperative programming language C.

Keywords

Residual Program Back Propagation Partial Evaluation Learning Pattern Automatic Specialization 
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]
    L.O. Andersen. C program specialization. Technical Report 92/14, DIKU, University of Copenhagen, Denmark, May 1992.Google Scholar
  2. [2]
    L.O. Andersen. Partial evaluation of C and automatic compiler generation (extended abstract). In U. Kastens and P. Pfahler, editors, Compiler Construction, Paderborn, Germany, October 1992. (Lecture Notes in Computer Science, vol. 641), pages 251–257. Springer-Verlag, 1992.Google Scholar
  3. [3]
    L. Beckman et al. A partial evaluator, and its use as a programming tool. Artificial Intelligence, 7(4): 319–357, 1976.MATHCrossRefGoogle Scholar
  4. [4]
    M.A. Bulyonkov. Polyvariant mixed computation for analyzer programs. Acta Information, 21: 473–484, 1984.MathSciNetMATHCrossRefGoogle Scholar
  5. [5]
    C.K. Gomard and N.D. Jones. Compiler generation by partial evaluation: a case study. Structured Programming, 12: 123–144, 1991.Google Scholar
  6. [6]
    H.F. Jacobsen. Speeding up the back-propagation algorithm by partial evaluation. DIKU Student Project 90-10-13, 32 pages. DIKU, University of Copenhagen. (In Danish), October 1990.Google Scholar
  7. [7]
    N.D. Jones, C.K. Gomard, and P. Sestoft. Partial Evaluation and Automatic Program Generation. Prentice-Hall, 1993. To appear.Google Scholar
  8. [8]
    N.D. Jones, P. Sestoft, and H. Søndergaard. Mix: A self-applicable partial evaluator for experiments in compiler generation. Lisp and Symbolic Computation, 2(1): 9–50, 1989.CrossRefGoogle Scholar

Copyright information

© Springer-Verlag/Wien 1993

Authors and Affiliations

  • Henrik Friborg Jacobsen
    • 1
  • Carsten Krogh Gomard
    • 2
  • Peter Sestoft
    • 3
  1. 1.DIKU, Department of Computer ScienceUniversity of CopenhagenCopenhagen ØDenmark
  2. 2.Computer Resources InternationalBirkerødDenmark
  3. 3.Department of Computer ScienceTechnical University of DenmarkLyngbyDenmark

Personalised recommendations