A Matrix Algebra for Neural Nets

  • Paul Cull
Part of the NATO Conference Series book series (NATOCS, volume 5)


Almost thirty-five years ago in their classic paper, McCulloch and Pitts (1943) described a method for modeling the nervous system. The basic idea of McCulloch and Pitts’ paper is that the nervous system can be described as a finite set of elements, called neurons, that have only two states, “on” and “off,” They assumed that time could be quantized into a set of discrete instants, so that the state of a neuron at the next instant of time would be a function of the present states of the neurons (and external inputs) that impinged on it.


Fast Fourier Transform Finite Field Matrix Algebra Convolution Theorem Field Linearization 
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.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    E.R. Caianiello, “Some remarks on the temorial linearization of general and linearly separable Boolean functions.” Kybernetik 12, 1973, pp. 90–93.CrossRefGoogle Scholar
  2. 2.
    P. Call, “Linear analysis of switching nets,” Kybernetik 8, 1971, pp. 31–39.CrossRefGoogle Scholar
  3. 3.
    J.L.S. da Fonseca and W.S. McCulloch, “Synthesis and linearization of nonlinear feedback shift registers.” MIT Research Laboratory of Electronics Quarterly Progress Report No, 86, 1967, pp. 355–366.Google Scholar
  4. 4.
    H.D. Landahl and R Runge, “Outline of a matrix calculus for neural nets.” Bull. Math. Biophys. 8, 1946, pp. 75–81,CrossRefGoogle Scholar
  5. 5.
    W.S. McCulloch and W.H. Pitts, “A logical calculus of the ideas immanent in nervous activity.” Bull. Math. Biophys. 5, 1943, pp. 115–133.CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media New York 1978

Authors and Affiliations

  • Paul Cull
    • 1
  1. 1.Department of Computer ScienceOregon State UniversityCorvallisUSA

Personalised recommendations