Outline of a Linear Neural Network and Applications
The aim of this paper is to show that additional perspectives are added to research on linear Neural Nets by utilizing a new definition of product, recently introduced (Caianiello, 92) in the context of Neural Nets defined on semirings rather than number fields. This we do by exhibiting its use for the design “on inspection” of a Neural Net which stores as one-step attractor any number of given patterns, with input basins of exactly specified tolerance. The paper is organized in the following way. In the next section we specify the notation and define the Neural Net model. In section three we describe applications in structured and not structured domains.
Unable to display preview. Download preview PDF.
- • Caianiello, E.R. (1992) Do All Neurons Know Algebra? Art. Neur. Networks II, I. Alexander and J. Taylor ed.s, 1, 221–227, Elsevier.Google Scholar
- • Duda, R.O., Hart, P.E. (1973) Pattern Classification and Scene Analysis, New York: J. Wiley and Sons.Google Scholar
- • Esposito, A., Rampone, S., Tagliaferri, R. (1992) A Neural Network for Error Correcting Decoding of Binary Linear Codes. IIASS tech. rep I-9901, in press on Neural Networks. Google Scholar
- • Gallager, R.G. (1968) Information Theory and Reliable Communication. New York: J.Wiley and Sons.Google Scholar
- • Huseyin, A. (1990) Vector Quantization. New York: IEEE Press.Google Scholar
- • MacWilliams, F.J., Sloane, N.J.A. (1981) The Theory of Error Correcting Codes. New York: North Holland.Google Scholar
- • McElice, R.J. (1977) The Theory of Information and Coding. In GianCarlo Rota ed., Encyclopedia of Mathematics and its Applications, S. London: Addison Wesley.Google Scholar
- • Peterson, W.W., Weldon, W.J. (1972) Error Correcting Code&. Cambridge: The MIT press.Google Scholar
- • Rampone, S. (1993) Linear Codes Interpolation from Noisy Patterns by Means of a Vector Quantization Process. IIASS Tech. Rep. I-9303, submitted for pubblication.Google Scholar
- • Shannon, C.E. (1959) Coding Theorem for a Discrete Source with a Fidelity Criterion. Ist. of Radio Eng., Int. Conv. Rec., 7, 142–163.Google Scholar