• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2004.10a
• /
• pp.145-148
• /
• 2004

규칙 기반 추론 시스템에서 규칙의 속성 감축은 다양한 방법으로 제안되어 왔다. 규칙의 속성 감축은 퍼지 추론 시스템을 구현하는데 있어서 처리 시간을 단축시킬 수 있으나 규칙의 종속성 및 상관성을 고려하지 않을 경우 예상하지 못한 추론 결과를 얻을 수 있다. 따라서, 본 논문에서는 복합속성을 가진 규칙의 속성 감축과 상관성을 고려하기 위하여 러프집합의 특성 중 식별가능 행렬과 식별가능 함수를 이용하였다. 그리고 속성 감축에 사용된 규칙은 복합속성(composite attribute)을 가지는 감성 데이터를 이용하였다.

• Journal of Internet Computing and Services
• /
• v.8 no.3
• /
• pp.109-118
• /
• 2007

This paper proposed a method that can be simply analyzed instead of the basic general Fuzzy rule that its insufficient characters are cut out. Based on the proposed method. Rough sets are used to eliminate the incomplete attributes included in the rule and also for a classification more precise; the agreement of the membership function's output extracted the maximum attributes. Besides, the proposed method in the simulation shows that in order to verify the validity, compare the max-product result of fuzzy before and after reducing rule hosed on the rice taste data; then, we can see that both the max-product result of fuzzy before and after reducing rule are exactly the same; for a verification more objective, we compared the defuzzificated real number section.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2007.11a
• /
• pp.261-264
• /
• 2007

퍼지 추론은 애매한 지식을 효과적으로 처리할 수 있는 장점이 있다. 그러나 규칙의 연관속성은 규칙을 과다하게 생성하기 때문에 유용하고 중요한 규칙을 결정하는데 여러 가지 문제점이었다. 본 논문에서는 퍼지 규칙에서 규칙간의 상관성을 고려하여 불필요한 속성을 제거하고, 퍼지규칙의 상대농도를 이용하여 추론결과의 정확성을 유지하면서 규칙의 수를 최소화 하는 방법을 제안한다. 제안한 방법의 타당성을 검증하기 위하여 기존의 규칙 감축 방법에 따른 출론 결과와 비교 검증하였다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2013.01a
• /
• pp.291-294
• /
• 2013

연관 규칙 감축 기법은 대규모 데이터를 사용하는 Socially aware computing분야에서 매우 중요한 이슈이다. 본 논문에서는 수집된 각종 데이터들을 각 속성 기준에 따라 이진 변환한 후 가중치를 부여하고 논리식 감축 방법을 이용하여 신뢰성을 보장하는 규칙을 도출하는 새로운 데이터 감축 기법을 제안한다. 이는 컴퓨터 시뮬레이션 결과 기존의 방식들에 비해 지지도, 신뢰도, 규칙 감소율, 연관 규칙 추출 시간에 좋은 성능을 보였으며 이는 빠른 시간 내에 신뢰성 높은 대규모 데이터 처리가 필요한 Socially aware computing분야에 적합하다고 판단한다.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2001.12a
• /
• pp.37-40
• /
• 2001

본 논문에서는 전문가 시스템에서 지식에 대한 규칙을 감소시키기 위해 러프 집합을 이용한 지식 감축 방법을 제안한다. 또한, 속성 항을 클래스로 분류하여 각 클래스와 이웃하는 클래스의 항들을 비교하여 리덕트와 코어를 구하여 최소화하였다. 이러한 방법은 방대한 양의 규칙을 최소화함으로써 의사결정 시간을 단축시킬 수 있다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.19 no.2
• /
• pp.285-290
• /
• 2009

An association rule mining has been studied to find hidden data pattern in data mining. A realization of fast processing method have became a big issue because it treated a great number of transaction data. The time which is derived by association rule finding method geometrically increase according to a number of item included data. Accordingly, the process to reduce the number of rules is necessarily needed. We propose the T-algorithm that is efficient rule reduction algorithm. The T-algorithm can reduce effectively the number of association rules. Because that the T-algorithm compares transaction data item with binary format. And improves a support and a confidence between items. The performance of the proposed T-algorithm is evaluated from a simulation.

• The Transactions of the Korea Information Processing Society
• /
• v.4 no.3
• /
• pp.846-858
• /
• 1997

Despite elaegant semantics and a lot of features, pure functional programming languages do not provide an affcient way of represnting states.Many researches have been done to resolve the problem, however, another problem arises that it is hard to implement becaese of the complex type system and redujction rule.Therefore, the scheme which simplifies the reduction rule and maintains states effciently is needed to have the implemen-taiton dffetive.This paper proposes st-calculus, the excution model of a functinal language with states and proves that the proposed model satistiies the church-Rosser theorem.It has simple reduction rules and the ability of rerpresenting states without, and the difficulties with implementation may be reduced by simplifving the reduction rules.

• Environmental and Resource Economics Review
• /
• v.31 no.4
• /
• pp.733-753
• /
• 2022

Korean government established the Nationally Determined Contribution (NDC) in 2015. After revising in 2019, the government updated an enhanced target at the end of last year. When the NDC is addressed, the emission targets of each sector, such as power generation, industry, and buildings, are also set. This paper analyzes the emission target of each sector by applying a claims problem or bankruptcy problem developed from cooperative game theory. The five allocation rules from a claims problem are introduced and the properties of each rule are considered axiomatically. This study applies the five rules on allocating carbon emission by sector under the NDC target and compares the results with the announced government target. For the power generation sector, the government target is set lower than the emissions allocated by the five rules. On the other hand, the government target for the industry sector is higher than the results of the five rules. In other sectors, the government's targets are similar to the results of the rule that allocates emissions in proportion to each claim.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.17 no.7
• /
• pp.881-886
• /
• 2007

Fuzzy inference has the advantage which can process the ambiguous knowledge. However the associated attributes of fuzzy rules are difficult to determine useful and important rules because the redundant attribute of rules is more than enough. In this paper, we propose a method to minimize the number of rules and preserve the accuracy of inference results by using fuzzy relative cardinality after removing unnecessary attributes from rough set. From the experimental results, we can see the fact that the proposed method provides better results (e.g the number of rules) than those of general rough set with the redundant attributes.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.9 no.4
• /
• pp.367-374
• /
• 1999

This paper suggests an integrated method based on concept ascension and attribute reduction for efficient induction of decision rules from a large database. We study an automatic scheme to generate concept trees by a clustering technique, a method for generalizing databases by the concept ascension technique, an optimal reduction method by means of attributes reduction using the sibmificance of attributes, and an efficient way of reduction of attribute values applying the discernible matrix and functions. The method can be used for the decision making tasks such as an investment planning or price evaluation, the construction of knowledge bases for diagnosis of defects or medical diagnosis, data analysis such as marketing or experimental data, information retrieval for high level inquiries, and so on.