• International Journal of Control, Automation, and Systems
• /
• v.6 no.6
• /
• pp.928-938
• /
• 2008

This study presents an approach for approximation an unknown function from a numerical data set based on the synthesis of a neuro-fuzzy model. An adaptive input data space parting method, which is used for building hyperbox-shaped clusters in the input data space, is proposed. Each data cluster is implemented here as a fuzzy set using a membership function MF with a hyperbox core that is constructed from a min vertex and a max vertex. The focus of interest in proposed approach is to increase degree of fit between characteristics of the given numerical data set and the established fuzzy sets used to approximate it. A new cutting procedure, named NCP, is proposed. The NCP is an adaptive cutting procedure using a pure function $\Psi$ and a penalty function $\tau$ for direction the input data space parting process. New algorithms named CSHL, HLM1 and HLM2 are presented. The first new algorithm, CSHL, built based on the cutting procedure NCP, is used to create hyperbox-shaped data clusters. The second and the third algorithm are used to establish adaptive neuro- fuzzy inference systems. A series of numerical experiments are performed to assess the efficiency of the proposed approach.

• Journal of KIISE:Software and Applications
• /
• v.33 no.8
• /
• pp.692-702
• /
• 2006

In this study, a weighted fuzzy min-max (WFMM) neural network model for pattern classification is proposed. The model has a modified structure of FMM neural network in which the weight concept is added to represent the frequency factor of feature values in a learning data set. First we present in this paper a new activation function of the network which is defined as a hyperbox membership function. Then we introduce a new learning algorithm for the model that consists of three kinds of processes: hyperbox creation/expansion, hyperbox overlap test, and hyperbox contraction. A weight adaptation rule considering the frequency factors is defined for the learning process. Finally we describe a feature analysis technique using the proposed model. Four kinds of relevance factors among feature values, feature types, hyperboxes and patterns classes are proposed to analyze relative importance of each feature in a given problem. Two types of practical applications, Fisher's Iris data and Cleveland medical data, have been used for the experiments. Through the experimental results, the effectiveness of the proposed method is discussed.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.1272-1277
• /
• 2003

This paper presents a method for combining classifier which is constructed by fuzzy and neural network classifiers and uses classifier fusion algorithms and selection algorithms. The input space of combing classifier is divided by the extended hyperbox region proposed in this paper to guarantee non-overlapped data property. To fuse the fuzzy classifier and the neural network classifier, we propose the fusion parameter for the overlapped data. In addition, the adaptive learning algorithm also proposed to maximize classifier performance. Finally, simulation examples are given to illustrate the effectiveness of the method.

• Proceedings of the KIEE Conference
• /
• 1999.11c
• /
• pp.636-639
• /
• 1999

본 논문에서는 기존의 FMMCNN이나 Fuzzy ART에서 Hyperbox를 정형으로 이용한 방법보다 적응적으로 분류가 가능한 컨벡스 집합을 기반으로 한 새로운 클래시피케이션 기법을 제안하였다. 컨벡스 다면체를 적응적으로 생성하기 위하여 퍼지 뉴럴 네트웍 분류기를 구성하고, 이를 이용한 패턴 클래스들을 생성하였다. 마지막으로, FMMCNN과의 다양한 시뮬레이션을 수행하여 본 논문의 우수성을 입증하였다.

• The KIPS Transactions:PartB
• /
• v.11B no.6
• /
• pp.717-722
• /
• 2004

This paper presents fuzzy rules to predict diagnosis of Wisconsin breast cancer using neural network with weighted fuzzy membership functions (NNWFM). NNWFM is capable of self-adapting weighted membership functions to enhance accuracy in prediction from the given 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 the enhanced bounded sums of n set of weighted fuzzy membership functions. Two number of prediction rules extracted from NNWFM outperforms to the current published results in number of rules and accuracy with 99.41%.

• Journal of KIISE:Software and Applications
• /
• v.32 no.1
• /
• pp.1-9
• /
• 2005

In this paper we propose a modified fuzzy min-max neural network model for pattern classification and discuss the usefulness of the model. We define a new hypercube membership function which has a weight factor to each of the feature within a hyperbox. The weight factor makes it possible to consider the degree of relevance of each feature to a class during the classification process. Based on the proposed model, a knowledge extraction method is presented. In this method, a list of relevant features for a given class is extracted from the trained network using the hyperbox membership functions and connection weights. Ft)r this purpose we define a Relevance Factor that represents a degree of relevance of a feature to the given class and a similarity measure between fuzzy membership functions of the hyperboxes. Experimental results for the proposed methods and discussions are presented for the evaluation of the effectiveness and feasibility of the proposed methods.

• KIPS Transactions on Software and Data Engineering
• /
• v.2 no.5
• /
• pp.341-346
• /
• 2013

In this paper, we propose a rule extraction method using a modified Fuzzy Min-Max (FMM) neural network. The suggested method supplements the hyperbox definition with a frequency factor of feature values in the learning data set. We have defined a relevance factor between features and pattern classes. The proposed model can solve the ambiguity problem without using the overlapping test process and the contraction process. The hyperbox membership function based on the fuzzy partitions is defined for each dimension of a pattern class. The weight values are trained by the feature range and the frequency of feature values. The excitatory features and the inhibitory features can be classified by the proposed method and they can be used for the rule generation process. From the experiments of sign language recognition, the proposed method is evaluated empirically.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2003.09a
• /
• pp.492-495
• /
• 2003

Fuzzy Min-Max Neural Network(FMMNN) is a powerful classifier, It has, however, some problems. Learning result depends on the presentation order of input data and the training parameter that limits the size of hyperbox. The latter problem affects the result seriously. In this paper, the new approach to alleviate that without loss of on-line learning ability is proposed. The committee machine is used to achieve the multi-resolution FMMNN. Each expert is a FMMNN with fixed training parameter. The advantages of small and large training parameters are used at the same time. The parameters are selected by performance and independence measures. The Decision of each expert is guided by the gating network. Therefore the regional and parametric divide and conquer scheme are used. Simulation shows that the proposed method has better classification performance.

• Journal of the Institute of Electronics Engineers of Korea TE
• /
• v.37 no.3
• /
• pp.114-122
• /
• 2000

This paper proposes a classification method using an amorphous Prototype to minimize classification error caused by such fixed-Prototype-based methods as Fuzzy C-Means, Nearest Neighborring Classification, FMMCNN, and Fuzzy-ART. For this method, a new fuzzy neural network is introduced, in which a convex polytope is generated or adaptively reshaped to classify the given datum into a proper group. Thus, this method contains a function to classify sequential data set. To show the validity of this method, various numerical experiments including comparison results with FMMCNN are presented

• Proceedings of the Korea Information Processing Society Conference
• /
• 2004.11a
• /
• pp.521-524
• /
• 2004

본 논문에서는 기존의 FMM 신경망의 학습 단계 중 하이퍼박스 축소 단계에서 발생하는 정보의 왜곡 현상을 보완하여, 패턴분류 성능을 향상시키는 방법론을 제안하고 그 유용성을 고찰한다. 제안된 학습방법에서는 기존 FMM 신경망의 하이퍼박스 생성, 확장, 축소의 3 가지 학습 과정 중 축소과정을 생략하고, 각 특징 범위와 빈도를 고려한 학습으로 대치함으로써 축소과정에서 나타나는 부작용을 보완할 수 있도록 하였다. 이는 가중치 개념을 통해 각 특징값의 상대적 중요도를 서로 다른 값으로 반영할 수 있도록 하며 하이퍼박스 중첩영역에서 발생하는 분류 오류를 최소화한다. 본 연구에서는 개선된 학습기법의 이론을 제시하고, 이를 아이리스 데이터 분류 실험에 적용하여 그 유용성을 고찰하였다.