Parallel computations and committee constructions
- 28 Downloads
The paper reviewed the results bearing out the deep-seated relation between the parallel computations and learning procedures for the laminated neural networks one of whose formalizations is represented by the theory of committee constructions. Additionally, consideration was given to two combinatorial problems concerned with learning pattern recognition in the class of affine committees—the problem of verifying existence of a three-element affine separating committee and that of element-minimal affine separating committee. The first problem was shown to be N P-complete, whereas the second problem is N P-hard and does not belong to the Apx class.
PACS numbers02.10.Ox 02.60.-x 89.20.Ff
Unable to display preview. Download preview PDF.
- 1.Mazurov, Vl.D., On the Committee of a System of Convex Inequalities, in Proc. Int. Math. Congr., Moscow: Mosk. Gos. Univ., 1966, no. 14, p. 41.Google Scholar
- 5.Judd, J.S., Neural Network Design and Complexity of Learning, New York: MIT Press, 1990.Google Scholar
- 6.Lin, J.H. and Vitter, J.S., Complexity Results on Learning by Neural Nets, Machine Learning, 1991, vol. 6, pp. 211–230.Google Scholar
- 11.Dinur, I., Regev, O., and Smyth, C., The Hardness of 3-uniform Hypergraph Coloring, in Proc. 43rd Ann. IEEE Sympos. Foundat. Comput. Sci., 2002.Google Scholar
- 12.Khachai, M.Yu., On the Computational and Approximational Complexity of the Problem of Minimal Separating Committee, Tavrich. Vest. Informatiki i Mat., 2006, no. 1, pp. 34–43.Google Scholar