뉴로-퍼지 가중치 제어 신경망 구조
The Neural Network Architecture with Neuro-Fuzzy Weighted-control
- 주제(KDC) 040.000
- 발행기관 東新大學校
- 발행년도 1995
- 총서유형 Journal
- 본문언어 한국어
초록/요약
In the learning of the neural network, Learning rate largely depends on the algorithm of network implementation, and is closely related to the weight length and computing speed. In this paper, we propose an algorithm to design digital logic systems and hardware using weighted control neural network with neuro-fuzzy architecture. The network for weight-control based on the mathmetical proposition has been proposed. The proposition use AT MOST n, AT LEAST n. The proposed algorithm has been applied to design JK flip-flop and ROM. The simulations shows that the proposed algorithm can simplify design of hardware, improve processing speed of system and implement synapse with good learning in the large neural network.
more목차
Ⅰ.서론
Ⅱ.가중치 제어 신경망
1.수학적 명제를 이용한 단충 신경망
2.가중치 제어 신경망
Ⅲ.뉴로-퍼지 가중치 제어 신경망
1.이진 논리와 퍼지 논리에 의한 XOR 신경망 구조
2.뉴로-퍼지 가중치 제어 신경망
Ⅳ.시뮬레이션
Ⅴ.결론