Abstract | ||
---|---|---|
This paper considers the classification properties of two-layer networks of McCulloch–Pitts units from a theoretical point of view. In particular we consider their ability to realise exactly, as opposed to approximate, bounded decision regions in R 2 . The main result shows that a two-layer network can realise exactly any finite union of bounded polyhedra in R 2 whose bounding lines lie in general position, except for some well-characterised exceptions. The exceptions are those unions whose boundaries contain a line which is “inconsistent,” as described in the text. Some of the results are valid for R n , n ⩾2, and the problem of generalising the main result to higher-dimensional situations is discussed. |
Year | DOI | Venue |
---|---|---|
1996 | 10.1006/jcss.1996.0026 | J. Comput. Syst. Sci. |
Keywords | Field | DocType |
exact classification,two-layer neural net,neural net | Discrete mathematics,Combinatorics,General position,Polyhedron,Hyperplane,Artificial neural network,Probability density function,Mathematics,Bounded function,Bounding overwatch | Journal |
Volume | Issue | ISSN |
52 | 2 | Journal of Computer and System Sciences |
Citations | PageRank | References |
4 | 0.47 | 6 |
Authors | ||
1 |
Name | Order | Citations | PageRank |
---|---|---|---|
Gavin J Gibson | 1 | 51 | 8.95 |