Title :
Studying the effects of multiplication neurons for parity problem
Author :
Li, Dazi ; Hirasawa, Kotaro ; Hu, Jinglu ; Murata, Junichi
Author_Institution :
Dept. of Electr. & Electron. Syst. Eng., Kyushu Univ., Fukuoka, Japan
Abstract :
It is known that the N-bit parity problem can be solved by a 3-layer sigmoidal neural network employing N/2 + 1 hidden units if N is even and (N + 1)/2 hidden units if N is odd. A particular type of multiplication unit is proposed. It is shown that networks constructed with one multiplication hidden unit can solve the N-bit parity problem easily.
Keywords :
arithmetic; neural nets; 3-layer sigmoidal neural network; N-bit parity problem; hidden units; multiplication neurons; multiplication unit; parity problem; Biological neural networks; Control systems; Feedforward neural networks; Feedforward systems; Humans; Neural networks; Neurons; Pattern recognition; Systems engineering and theory; Testing;
Conference_Titel :
SICE 2002. Proceedings of the 41st SICE Annual Conference
Print_ISBN :
0-7803-7631-5
DOI :
10.1109/SICE.2002.1195515
