• 한국지능시스템학회논문지
• /
• 제1권1호
• /
• pp.9-25
• /
• 1991

This tutorial paper has been written for biologists, physicians or beginners in fuzzy sets theory and applications. This field is introduced in the framework of medical diagnosis problems. The paper describes and illustrates with practical examples, a general methodology of special interest in the processing of borderline cases, that allows a graded assignment of diagnoses to patients. A pattern of medical knowledge consists of a tableau with linguistic entries or of fuzzy propositions. Relationships between symptoms and diagnoses are interpreted as labels of fuzzy sets. It is shown how possibility measures (soft matching) can be used and combined to derive diagnoses after measurements on collected data. The concepts and methods are illustrated in a biomedical application on inflammatory protein variations. In the case of poor diagnostic classifications, it is introduced appropriate ponderations, acting on the characterizations of proteins, in order to decrease their relative influence. As a consequence, when pattern matching is achieved, the final ranking of inflammatory syndromes assigned to a given patient might change to better fit the actual classification. Defuzzification of results (i.e. diagnostic groups assigned to patients) is performed as a non fuzzy sets partition issued from a "separating power", and not as the center of gravity method commonly employed in fuzzy control. It is then introduced a model of fuzzy connectionist expert system, in which an artificial neural network is designed to build the knowledge base of an expert system, from training examples (this model can also be used for specifications of rules in fuzzy logic control). Two types of weights are associated with the connections: primary linguistic weights, interpreted as labels of fuzzy sets, and secondary numerical weights. Cell activation is computed through MIN-MAX fuzzy equations of the weights. Learning consists in finding the (numerical) weights and the network topology. This feed forward network is described and illustrated in the same biomedical domain as in the first part.

• Nuclear Engineering and Technology
• /
• 제53권11호
• /
• pp.3675-3684
• /
• 2021

Human reliability analysis (HRA) is a proactive approach to model and evaluate human systematic errors, and has been extensively applied in various complicated systems. Dependence assessment among human errors plays a key role in the HRA, which relies heavily on the knowledge and experience of experts in real-world cases. Moreover, there are ofthen different types of uncertainty when experts use linguistic labels to evaluate the dependencies between human failure events. In this context, this paper aims to develop a new method based on linguistic hesitant fuzzy sets and the technique for human error rate prediction (THERP) technique to manage the dependence in HRA. This method handles the linguistic assessments given by experts according to the linguistic hesitant fuzzy sets, determines the weights of influential factors by an extended best-worst method, and confirms the degree of dependence between successive actions based on the THERP method. Finally, the effectiveness and practicality of the presented linguistic hesitant fuzzy THERP method are demonstrated through an empirical healthcare dependence analysis.

• 전자공학회논문지C
• /
• 제34C권7호
• /
• pp.61-69
• /
• 1997

With the vigorous researches in the character recognition, the need to recognize run-on multilingual handwritten characters is increasing to provide uses with more comfortable PUI(pen user interface) environments. In general, many intermediate word candidates word candidates are generated in run-on multilingual recognition because there is no information of ending position and linguistic kind of character. To remove unnecessary word candidates which are generated in run-on multilingual recognition, we classify them into two groups and select the best candidate among the word candidates in the group where the final characater is completed using 5 attributes. In this research, we propose a method in order to select the best one candidate. It is called WRM (Weighted ranking method). The weights are adaptively trained by LMS(Least mean square) learning rule. Results show that the abilities of decision makin gusing weights are much better than those not using weights.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 1998년도 The Third Asian Fuzzy Systems Symposium
• /
• pp.404-409
• /
• 1998

In this paper, we propose a learning algorithm of fuzzy neural networks with trapezoidal fuzzy weights. This fuzzy neural networks can use fuzzy numbers as well as real numbers, and represent linguistic information better than standard neural networks. We construct trapezodal fuzzy weights by the composition of two triangles, and devise a learning algorithm using the two triangular membership functions, The results of computer simulations on numerical data show that the fuzzy neural networks have high fitting ability for target output.

• 한국지능시스템학회논문지
• /
• 제10권3호
• /
• pp.218-227
• /
• 2000

본 논문에서는 시소러스에 근거한 새로운 퍼지 정보검색 기법을 제안한다. 제안된 방법에서 시소러스는 내부 용어들 간의 관련도를 정성적인 언어 갑으로 갖는 퍼지 언어 매트릭스로 표현되며 용어들간의 관계는 동의, 계층, 그릭 연관이 세 가지 관계가 제공된다 싯러스 내부 용어들 간이 무시된 관련도가 퍼기 이론에 근거한 퍼지이론에 근거한 퍼지 언어 매트릭스의 전이 폐쇄 알고리즘에 의해 추론된다 또한 제안돈 방법은 사용자의 질의, 그리고 문서와 같은 정보 항목의 표현에도 인간이 주관적이고 부정확한 측도를 그대로 반영하는 정성적인 언어 값을 허용한다. 따라서 논문 [1-3]에서 제안된 방법보다 좀 더 유용하다. 또한 질의 평가시 퍼지 언어 매트릭스와 AON(Associated Ordinary Number)값을 이용하기 때문에 논문 [1-3]에서 사용되는 방법보다 시간적으로 효츌적이다. 결과적으로 사용자가 좀 더 유용하고 지능적인 방법으로 질의를 처리할 수있도록 한다

• 대한인간공학회:학술대회논문집
• /
• 대한인간공학회 1995년도 춘계학술대회논문집
• /
• pp.25-29
• /
• 1995

When we design a product, we need to consider human sensibility for the product. In this study, we developed a technique to measure human sensibility for a product. Because human sensibility for a product is very subjective and fuzzy, it is hard to measure easily. To deal with this difficulty effectively, we used fuzzy-weighted checklist to this problem. The fuzzy- weighted checklist presents a fuzzy version of the weighted checklist technique computerized for evaluating or comparing complex system (or subjects). In this technique, we used pairwise comparison to get the relative weights of wensibility factors. Also, we used linguistic ratings to get the scores of sensibility factors for a product. Then, we synthesize the scores of sensi- bility factors to get fuzzy composite score (and linguistic approximation). If there are several alternatives, we can conduct alternative comparison. Finally, we wrote the program of this technique by Quick Basic software. As an example, this technique applied to car. The results show that we can use this technique effectively to the quantitative evaluation of human sensibility

• 대한인간공학회지
• /
• 제15권1호
• /
• pp.15-26
• /
• 1996

When a product is designed, it is important to consider its image on consumers. In this study, we developed a technique to measure product image. Because human image of a product is very subjective and fuzzy, it is difficult to measure easily. To deal with this difficulty effectively, we used fuzzy- weighted checklist. The fuzzy-weighted checklist presents a fuzzy version of the weighted checklist technique for evaluating or comparing complex systems or subjects. In this technique, we used a pairwise comparison method to obtain the relative importance weights of image factors. Also, we used linguistic ratings to obtain the scores of image factors for a product. Then, we synthesized the scores of image factors to obtain a fuzzy composite score and its linguistic approximation. The entire procedure of this technique was written in quick Basic. As an example, this techinque is applied to car evaluation. The results show that this technique can be effectively used to the quqntitative evaluation of huamn image.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제2권3호
• /
• pp.210-214
• /
• 2002

Association rule mining is an exploratory learning task to discover some hidden dependency relationships among items in transaction data. Quantitative association rules denote association rules with both categorical and quantitative attributes. There have been several works on quantitative association rule mining such as the application of fuzzy techniques to quantitative association rule mining, the generalized association rule mining for quantitative association rules, and importance weight incorporation into association rule mining fer taking into account the users interest. This paper introduces a new method for generalized fuzzy quantitative association rule mining with importance weights. The method uses fuzzy concept hierarchies fer categorical attributes and generalization hierarchies of fuzzy linguistic terms fur quantitative attributes. It enables the users to flexibly perform the association rule mining by controlling the generalization levels for attributes and the importance weights f3r attributes.

• 한국건설관리학회논문집
• /
• 제5권2호
• /
• pp.115-122
• /
• 2004

이 논문은 건설공사 과정에서 불확실성과 위험성이 비교적 높은 토공사, 지정공사 및 기초공사에서 발생할 수 있는 리스크인자의 중요도 산정에 관한 것이다. 이 연구는 리스크 $식별\cdot분석\cdot대응$으로 이루어지는 리스크관리 3단계 중 리스크 식별과 분석단계를 중심으로 연구를 진행하였다. 리스크 식별은 기존의 건설공사 작업분류체계를 참고하여 대상 공종을 $공통\cdot토공사\cdot지정$ 및 기초공사로 구분하여 초기 건설공사의 리스크 분류체계를 제시하였다. 리스크 분석은 리스크분류체계를 바탕으로 퍼지이론에 기반하여 실시하였다. 리스크인자의 중요도는 AHP기법에 의한 상대적 중요도와 퍼지척도로부터 구한 리스크인자들 사이의 절대적 중요도를 고려하여 산정하였으며 리스크 인자의 최종적인 중요도는 Sugeno $\lambda$-퍼지척도를 사용하여 구하였다.

• 한국정보전자통신기술학회논문지
• /
• 제9권1호
• /
• pp.75-81
• /
• 2016

소설에서 주요 인물은 소설의 이야기를 전개하는 아주 중요한 역할을 담당하여 소설에서 없어서는 안 되는 중심인물을 의미한다. 기존의 인물명 인식 연구에서는 구축해놓은 인물명 사전을 통해 인물명을 인식하였고, 영어의 경우 대소문자 구별이 있으며 인물명과 함께 사용되는 단어를 활용하여 인물명을 인식하였다. 본 논문에서는 한국어 소설에서 용언, 규칙 및 가중치를 이용한 주요 인물명 인식 기법에 대해 제안한다. 먼저, 인물이 행할 수 있는 용언을 근거로 인물명 후보를 인식하고, 인식된 인물명 후보 중 인물명으로 사용될 수 없는 규칙에 해당되는 후보들을 제거한다. 문장에 나타나는 인물명 후보의 수에 따라 가중치를 부여하여 중요도를 계산하고, 중요도가 임계치 이상인 경우 주요 인물명으로 판단한다. 소설 300권을 대상으로 실험 결과 평균 85.97%의 정확도를 보였다. 인식된 주요 인물명은 향후 소설내 등장인물 간 연관관계를 파악하거나 등장인물의 행위, 성향 등을 파악하는데 활용될 수 있다.