This Article 
 Bibliographic References 
 Add to: 
Geometrical Learning Algorithm for Multilayer Neural Networks in a Binary Field
August 1993 (vol. 42 no. 8)
pp. 988-992

A geometrical expansion learning algorithm for multilayer neural networks using unipolar binary neurons with integer connection weights, which guarantees convergence for any Boolean function, is introduced. Neurons in the hidden layer develop as necessary without supervision. In addition, the computational amount is much less than that of the backpropagation algorithm.

[1] J. Hopfield and D. Tank, "Neural computation of decisions in optimization problems,"Biolog. Cybern., vol. 52, pp. 141-152, 1985.
[2] J. McClelland and D. Rumelhart,Explorations in Parallel Distributed Processing. Cambridge, MA: MIT Press, 1988.
[3] C. Mead and M. Ismail,Analog VLSI Implementation of Neural Systems. Kluwer Academic Publishers, 1989.
[4] S. Muroga,Threshold Logic and Its Applications. New York: Wiley, 1971, p. 273.
[5] S. Ghosh and A. K. Choudhury, "Partition of Boolean functions for realization with multithreshold logic elements,"IEEE Trans. Comput., vol. C-22, no. 2, Feb. 1973.

Index Terms:
geometrical learning algorithm; multilayer neural networks; binary field; unipolar binary neurons; integer connection weights; Boolean function; hidden layer; Boolean functions; feedforward neural nets; learning (artificial intelligence).
S.K. Park, J.H. Kim, "Geometrical Learning Algorithm for Multilayer Neural Networks in a Binary Field," IEEE Transactions on Computers, vol. 42, no. 8, pp. 988-992, Aug. 1993, doi:10.1109/12.238491
Usage of this product signifies your acceptance of the Terms of Use.