퍼지신경망과 비중복면적 분산 측정법을 이용한 최소의 특징입력 및 퍼지규칙의 추출

국문 초록 (Abstract)

본 논문은 가중 퍼지소속함수 기반 신경망(Neural Network with Weighted Fuzzy Membership Functions, NEWFM)을 이용하여 위스콘신 유방암(Wisconsin breast cancer)의 진단을 수행하는 퍼지규칙을 추출하고 비중복면적 분산 측정법을 사용하여 특정입력수를 최소로하는 방안을 제안하고 있다. NEWFM 구조의 중간 부분인 하이퍼박스(hyperbox)들은 n 개의 대, 중, 소로 구성된 가중 퍼지소속함수 집합으로 구성되며, 학습 후 각 집합의 대, 중, 소로 구성된 가중 퍼지소속함수는 퍼지집합의 경계합(bounded sum)을 사용하여 다시 하나의 가중 퍼지소속함수로 합성(BSWFM) 된다. n 개의 특정입력 (feature input)은 학습된 모든 하이퍼박스에 연결되어 예측 작업을 수행한다. 여기에 비중복면적 분산 측정법을 적용하여 중요도가 낮은 특정입력을 제거하면서 최소의 m 개 특정업력만을 사용한 하이퍼박스로 단순화 시킨다. 이러한 방법으로 위스콘신 유방암의 9개의 특징입력 중 4개를 사용하여 NEWFM으로 추출된 2개의 퍼지규칙은 99.71%의 예측 인식율을 가지며 이는 퍼지규칙의 수와 인식율에 있어 현재 발표된 논문의 결과보다 우수함을 보여준다.

다국어 초록 (Multilingual Abstract)

This paper presents fuzzy rules to predict diagnosis of Wisconsin breast cancer with minimized number of feature input using the neural network with weighted fuzzy membership functions (NEWFM) and the non-overlap area distribution measurement method. NEWFM is capable of self-adapting weighted membership functions from the given the Wisconsin breast cancer clinical training data. n set of small, medium, and large weighted triangular membership functions in a hyperbox are used for representing n set of featured input. The membership functions are randomly distributed and weighted initially, and then their positions and weights are adjusted during learning. After learning, prediction rules are extracted directly from n set of enhanced bounded sums of n set of small, medium, and large weighted fuzzy membership functions. Then, the non-overlap area distribution measurement method is applied to select important features by deleting less important features. Two sets of prediction rules extracted from NEWFM using the selected 4 input features out of 9 features outperform to the current published results in number of set of rules, number of input features, and accuracy with 99.71%.

목차 (Table of Contents)

• 요약
• Abstract
• 1. 서론
• 2. 가중 퍼지소속함수 기반 신경망 (Neural Network with Weighted Fuzzy Membership Function, NEWFM)
• 3. 실험 결과 (Experimental Results)
• 4. 결론
• 참고문헌
• 저자소개

참고문헌 (Reference)

1 "Voting in Fuzzy Rule-Based Systems for Pattern Classification Problems" 103 : 223-238, 1999.

2 "Simplified Fuzzy ARTMAP" 19-25, 1993.

3 "Self-Adaptive Neuro-Fuzzy Inference System for Classification Applications" 10 (10): 790-802, 2002.

4 "Neural-network-based fuzzy logic control and decision system" 40 (40): dec.1991.

5 "Multisurface Method of Pattern for Medical Diagnosis Applied to Breast Cytology Proc. National Academy of Sciences" 87 : 9193-9166, 1990.

6 "Modeling and Control of Carbon Monoxide Concentration Using a Neuro-Fuzzy technique IEEE Trans." 3 : 271-279, 1995.

7 "Fuzzy min-max neural networks-Part 1：Classification" neural networks (neural networks): 776-786, 1992.

8 "Fuzzy Rules to Predict Degree of Malignancy in Brain Glioma" 40 : 2002.

9 "Foundation of Neural Networks The MIT Press" 1996.

10 "Finding Fuzzy Rules by Neural Network with Weighted Fuzzy Membership Function International Journal of Fuzzy Logic and Intelligent Systems" 4 (4): 211-216, 2004.

11 "Approximative Fuzzy Rules Approaches for Classification with Hybrid-GA Techniques" 136 : 193-214, 2001

12 "An On-Line Self-Constructing Neural Fuzzy Inference Network and Its Applications" 6 (6): 12-32, 1998.

13 "ARTMAP： Supervised real-time learning and classification of nonstationary data by a self-organizing neural network" 4 : 565-588, 1991.

14 "A Neuro-Fuzzy Method to Learn Fuzzy Classification Rules from Data" 89 : 277-288, 1997.

15 "A Neural Networks with Disjunctive Fuzzy Information" 11 : 1113-1125, 1998.

16 "A Modified Fuzzy Min-Max Neural Network for Pattern Classification" Springer-Verlag 792-798, 2004