• 한국게임학회 논문지
• /
• 제9권6호
• /
• pp.117-125
• /
• 2009

본 논문은 포석 바둑을 위해, 패턴 지식을 근간으로 바둑 용어 지식을 수행할 수 있는 뉴로-퍼지 추론에 대한 실험 결과를 설명하였다. 즉, 포석 시 최선의 착점을 결정하기 위한 뉴로-퍼지 추론 시스템의 구현을 논하였다. 또한 추론 시스템의 성능을 시험하기 위하여 시차 학습(TD($\lambda$) learning) 시스템과의 대결을 벌였다. 대결 결과에 의하면 단순한 뉴로-퍼지 추론 시스템조차 시차 학습 모델과 충분히 대결할 만하며, 뉴로-퍼지 추론 시스템이 실제 바둑 게임에도 적용될 수 있는 잠재력을 보였다.

• 한국지구과학회지
• /
• 제23권2호
• /
• pp.154-165
• /
• 2002

The purpose of this study was to examine the effects of metastrategic exercise on a scientific reasoning strategy to control variables, and investigate the developmental patterns in the strategy usage within a given period. Two groups composed of 90 fifth grade students engaged in a scientific reasoning task over six daily sessions. Additionally, one group engaged in metastrategic exercise on fictional students' strategies of controlling variables on the task, while the other spent equivalent time on an unrelated task. Based upon results of the study, the following conclusions can be drawn. First, the metacognitive exercise on the strategy to control variables has positive and long-standing effects on the strategy performance at the reasoning task. The exercise also takes effect of near-transfer. Taking into consideration only about sixty minutes of metastrategic practice, the results provide the validity of the activity in order to develop children's reasoning strategies. Second, in a scientific reasoning task, each child seems to go through one out of two developmental patterns in their usage of reasoning strategies: gradual change or fundamental change. Considering the ratio of pattern of fundamental change between the two groups, it is clear that the metacognitive exercise influences the developmental pattern of strategy usage.

• 한국지능시스템학회논문지
• /
• 제4권3호
• /
• pp.32-48
• /
• 1994

This paper proposes the compensation algorithm of conventional shift pattern using fuzzy reasoning in automatic transmission vehicles. Recently, automatic transimssion vehicles are continually increasing because of theire ease to drive. Also users require the high performance which includes the smooth shift quality and shift scheduling that matches driver;s intentions. So the shift scheduling has been inproved significantly through the application of electronic control. But, in spite of this development, vehicles using conventional shift pattern are sometimes in discord with driver's intention on roads. In this paper, the paper, the proposed compensation algorithm makes a automatic transmission vehicle be able to select an optimal gear shifting time and position using fuzzy reasoning and make a vehicle driver feel confortable even when the vehicle runs on roads which is extremely changed. Therefore, a vehicle driver can expect to reduce the nimber of unnecessary gear shifting and expect the fuel efficiency high. To show usefulness of the proposed method, some simulation are made to compared with conventional gear shifting. Paper prosposes the compensation mehtod of conventional shift pattern using fuzzy reasoning for the purpose that a vehicle can select an optimal gerar shifting time and position in automatic vehicle. Though the conventional shift pattern has no pliability, vehicle driver can feel comfortable and can reduce the number of unnecessary gear shifting using the proposed method on variable road condition. Therefore, it can be expected the fuel efficiency.

• 한국멀티미디어학회논문지
• /
• 제4권2호
• /
• pp.189-196
• /
• 2001

퍼지 논리의 추론과정에서 일부의 정보가 무시되어 적절하지 못한 추론 결과를 초래 할 수 있다. 한편 신경망은 패턴 처리에는 적합하지만 인간의 지식을 모델링하기 위해서 필요한 논리적인 추론에는 부적합하다. 그러나 신경망의 변형인 신경 논리망을 이용하면 논리적인 추론이 가능하다. 따라서 본 논문에서는 기존의 신경 논리망을 기반으로 하는 추론네트워크를 확장하여 퍼지 추론 네트워크를 구성한다. 그리고 기존의 추론 네트워크에서 사용되는 전파규칙을 보완하여 적용한다. 퍼지 추론 네트워크상에서 퍼지 규칙의 실행부에 해당하는 명제의 믿음 값을 결정하기 위해서는 추론하고자 하는 명제에 연결된 노드들을 탐색해야 한다.

• 한국산업정보학회:학술대회논문집
• /
• 한국산업정보학회 2000년도 추계공동학술대회논문집
• /
• pp.375-389
• /
• 2000

Fuzzy logic ignores some information in the reasoning process. Neural networks are powerful tools for the pattern processing, but, not appropriate for the logical reasoning. To model human knowledge, besides pattern processing capability, the logical reasoning capability is equally important. Another new neural network called neural logic network is able to do the logical reasoning. Because the fuzzy inference is a fuzzy logical reasoning, we construct fuzzy inference network based on the neural logic network, extending the existing rule- inference. network. And the traditional propagation rule is modified. For the search strategies to find out the belief value of a conclusion in the fuzzy inference network, we conduct a simulation to evaluate the search costs for searching sequentially and searching by means of search priorities.

• 한국지구과학회지
• /
• 제35권5호
• /
• pp.385-398
• /
• 2014

This study investigated students' alternative conceptions of plate boundaries and their conception revision according to the pattern of students' reasoning. Participants were twenty-two 10th grade high school students. All participants were asked to draw the three types of plate boundaries and to explain their drawings. Nine students participated in the reasoning activity. To this end, a semi-structured interview was conducted during which key questions were asked for the students to individually answer. The key questions used in the reasoning activity were created, by utilizing questions used in the previous studies. The findings revealed that the alternative conceptions of plate boundaries were classified into three levels based on established criteria. Students who attempted a variety of reasoning strategies such as causal reasoning, using an analogy, abductive reasoning, data reconstruction and concept combination, revised their alternative conception to a scientific conception after the reasoning activity. On the other hand, some students could not revise their alternative conceptions because they only conducted an incomplete reasoning strategy. The study also found that they were unable to use other reasoning strategies, either.

• 한국초등과학교육학회지:초등과학교육
• /
• 제23권1호
• /
• pp.61-73
• /
• 2004

본 연구는 다양한 경향성을 발견할 수 있는 과제를 피험자들에게 제시하고, 피험자들이 경향성 지식을 생성하는 과정에서 표상 된 지식의 종류와 사고 유형을 분석하여 경향성 지식의 생성 과정을 알아보고자 하였다. 본 연구의 결과, 경향성 지식의 생성 과정에서 표상 된 지식은 요소, 요소변화, 관련선지식, 예측경향성, 최종 경향성 지식 등 5가지 유형의 과정적 지식이 생성되었다. 그리고 사용된 사고 유형은 `대상인식', '관련지식회상', '요소 또는 요소변화의 탐색', '예측경향성 발견', '예측경향성 확인', 경향성 조합', '경향성 선택' 등 7가지 유형의 사고가 경향성 지식 생성 과정에 관여함을 볼 수 있었다. 또한 경향성 지식 생성과정은 `요소 \$\longrightarrow$요소변화 \$\longrightarrow$(관련선지식 \$\longrightarrow$)예측경향성 생성ㆍ확인 \$\longrightarrow$최종경향성'의 순으로 과정적 지식들이 단계적으로 표상 되어 생성되었으며, 이러한 과정에서 귀납적 추론뿐만 아니라, 귀추적 추론과 연역적 추론도 함께 관여하는 것을 볼 수 있었다.

• Asia pacific journal of information systems
• /
• 제14권3호
• /
• pp.77-92
• /
• 2004

When humans make decisions, they differentiate classifications of individual attribute variables that affect the decisions according to the importance and pattern of each attribute variables. The present study examines the practicality of the proposed Code Arrangement-Based Reasoning(CABR), which resembles the human's way of reasoning. To this end, we developed a CABR technique that classifies each attribute variable affecting significant impacts on the target variable into a cluster and assigns a code to the cluster. For verifying the proposed technique, both case-based reasoning and CABR were used for the customer continuance judgment problem of an automobile insurance company. Results indicated that the performance of CABR is close to the one of the case-based reasoning. The CABR also shows the possibility of using bio-informatics techniques for organizational data analysis in the future.

• 한국과학교육학회지
• /
• 제23권4호
• /
• pp.309-318
• /
• 2003

Scientific reasoning is one of the main concerns in current science education. This study have tried to answer on the question whether Korean science education has the potential to help improve of students' ability to think scientifically. Therefore, the present study investigated the relationship between reasoning patterns evident in science textbook and science examination items, and students' scientific reasoning skills across grades in Korea. 1975 subjects (1022 females and 953 males) were administered in the Lawson's Test of Scientific Reasoning skills. Forty seven science textbooks and 240 assessment instrument were analyzed by several scientific reasoning keys. Scientific reasoning patterns were adopted from Lawson's classification which characterized the patterns as the empirical-inductive and the hypothetical-deductive. This study found that reasoning patterns evident in textbook analyses and assessment instrumental items do not evidentce the potential to stimulate the development of students' reasoning skill. In order to improve the students' abilities to think and achieve, higher levels of reasoning must be included in the science textbook and examination. Further, some of scientific reasoning processes, such as generating hypotheses, designing experiments, and logical prediction, were not found in science textbooks and test items in Korean secondary schools. This study also discussed the educational implication of these results and further studies about to develop student's reasoning ability.

• 한국학교수학회논문집
• /
• 제20권4호
• /
• pp.405-422
• /
• 2017

Nim 게임을 구분하여 한 더미 대상 게임을 1단계, 두 더미 대상 게임을 2단계, 세 더미 대상 게임을 3단계로 나누어 중학교 수학영재들을 대상으로 탐구활동을 실시하였다. 학생들은 난이도가 낮은 1단계에서는 연역적 추론을 통하여 쉽게 필승전략을 발견하였다. 2단계에서는 연역적 추론 또는 귀납적 추론으로 필승전략을 발견하였지만 귀납적 추론 과정에서는 오류가 발견되었다. 3단계 게임에서는 연역적 추론으로 필승전략을 발견한 학생들은 없었으며 귀납적 추론 과정에서는 오류가 발견되었다. 유한개의 경우에서 성립하는 패턴을 정당화 절차 없이 무조건 일반화하려는 경향이 오류의 원인임이 밝혀졌다. 학생들에게 이진법 상자를 시각적으로 제시한 결과, 학생들은 승패에 따른 패턴을 쉽게 발견하고 게임 활동을 통하여 필승전략을 인식하게 되었으며 일부 학생들은 발견한 필승전략을 정당화하는 단계에 도달할 수 있었다.