상세정보
학술대회
A Study on Neural Network Recognizer based on Fuzzy Rules and Fuzzy Inference: Fuzzy Driven Neural Network Recognizer in Pattern Recognition
- 발행일
- 201404
- 출처
- International Conference on Information Science, Electronics and Electrical Engineering (ISEEE) 2014, pp.1913-1917
- 협약과제
- 13SE1600, 시각 생체 모방 소자 및 인지 시스템 기술 개발, 정명애
- 초록
- In this study, we introduce neural network recognizer based on fuzzy rules and fuzzy inference. The use of neural networks is proposed for efficient implementation of the fuzzy inference and the neural network is a trainable device consisting of some fuzzy rules and three processes, namely, premise, consequence and fuzzy inference processes. The premise process is driven by fuzzy c-means and the consequence processes deals with a polynomial function. A learning algorithm for the neural network recognizer is developed and its performance is compared with that of previous studies.
- KSP 제안 키워드
- Pattern recognition, efficient implementation, fuzzy C-Means, fuzzy inference, fuzzy rules, learning algorithms, neural network, polynomial function