Separating capacity of a cascaded network of threshold elements

作者: Ryuzo Takiyama

DOI: 10.1002/ECJA.4400661205

关键词:

摘要: This paper discusses the information processing ability of cascade network threshold elements, which is a fundamental structure to realize piecewise-linear machine. upper bounds for number dichotomies and separating capacity are determined that type network, where theory equivalent transformation from two-layered utilized positively. As result, it shown bound consisting k elements given by 2k(n + 1), n dimension patterns be classified.

参考文章(10)
Bitner, Capacity and Efficiency of Decision Functions IEEE Transactions on Computers. ,vol. 26, pp. 1147- 1151 ,(1977) , 10.1109/TC.1977.1674763
Ryuzo Takiyama, A learning procedure for multisurface method of pattern separation Pattern Recognition. ,vol. 12, pp. 75- 82 ,(1980) , 10.1016/0031-3203(80)90005-9
C.S. Weaver, Some Properties of Threshold Logic Unit Pattern Recognition Networks IEEE Transactions on Computers. ,vol. 24, pp. 290- 298 ,(1975) , 10.1109/T-C.1975.224209
S.-I. Amari, Learning Patterns and Pattern Sequences by Self-Organizing Nets of Threshold Elements IEEE Transactions on Computers. ,vol. C-21, pp. 1197- 1206 ,(1972) , 10.1109/T-C.1972.223477
Thomas M. Cover, Geometrical and Statistical Properties of Systems of Linear Inequalities with Applications in Pattern Recognition IEEE Transactions on Electronic Computers. ,vol. EC-14, pp. 326- 334 ,(1965) , 10.1109/PGEC.1965.264137
Lubomír Ohera, On the problem of separability of some pattern recognition learning systems. Kybernetika. ,vol. 5, pp. 420- 436 ,(1969)