컴퓨터 프로그램에 의한 의사결정분석

李昌孝 新羅大學校 1984 論文集 Vol.16 No.-

The aim of decision analysis is to try to supply some aids to the process of decision-making especially where the decision situation involves uncertainty. The idea is that, by proving the analytical framework, the manager will be able to concentrate on supplying the human inputs of intuition and experience one by one withut having to try to juggle these in his head, while at the same time tryinjg to provide further inputs. in this paer, the author concentrated on developing the computer program of Bayesian decision procedure with data and the procedure of computing the expected payoff with continuous variables. This computer program was written in Apple Ⅱ BASIC language and the author used the TriGem020 micro-computer with 48K memory.

多目的計劃法에서의 非劣位解 및 選好解 探索

李昌孝 新羅大學校 1985 論文集 Vol.18 No.-

The noninferior solutions searching methods(e, g., weighting method and e-constraint method) develop a representation of the noninferiour set for consideration by the decision maker (DM), and emphasize the development of information about a multiobjective problem that is presented to a DM in a manner that allows the range of choice and the trade-offs among objectives to be well understood. However, one disadvantage which has severely limited there method's practical applicability is that they usually generate a large number of noninferior solutions; it becomes almost impossible for the DM to choose one which is the most satisfactory, i, e. , a preferred solution among them. Therefore, the methods are generally incorporated into some of the preferred solution searching methods. Among the preferred solution searching methods two representative methods, namely, the multiattribute utility function (MAUF) method and the surrogate worth trade-off(SWT) method use global and local utility (preferrence) modelling, respective. However, the MAUF method developed by Keeney and Raiffa have critical disadvantage such that it may force the DM to fit a function not truly representing his or her preferences. On the other hand, the SWT method developed by Haimes and Hall based on local utility function modelling, provides an alternative approach that avoids the restrictive assumpt-ions. Instead of specifying the utility function globally, their procedures construct a sequence of local preference approximations of it. The SWT method uses the e-constraint problem as a means of generating noninferior solutions. Trade-offs among objectives, whose values are expressed by values of strictly positive Lagrangian multipliers, are used as a medium. The DM responds by expressing his degree of preference over the prescribed trade-offs and by assigning numerical values to each surrogate worth function. This method guarantees the generated solution in each iteration to be Parto optimal and the DM can select his preferred solution from among noninferior solutions. However, original version of the SWT method has many disadvantages. Therefore, in this paper, a new multiobjective decision making technique for solvlng multiobjective optimization problems: the Ready Surrogate Worth Trade-off(RSWT) method which belongs to revised SWT methods is presented by the author. Using this technique, the preferred solution for the DM can be described more conveniently and efficiently from among a Pareto optimal solution sat.

광음향 효과에 의한 고체의 열확산도 결정

이창효 漢陽大學校 基礎科學硏究所 1987 基礎科學論文集 Vol.6 No.-

Rosencwaig-Gersho 이론을 수치 해석함으로써 고체시료의 열확산도를 결정하는 새로운 방법을 알아내었다. 동일 두께의 고체시료에 대해서는 부착한 배경물질의 종류에 관계없이 광음향신호의 세기가 같아지는 광단속진동수가 존재하며, 이 광단속진동수는 시료의 두께에 따라 달라진다. 이러한 광단속진동수와 시료의 두께 사이 관계를 log-log 그래프로 표시하면 직선으로 나타나며 ?? 일 때의 직선과의 간격으로 열확산도를 결정할 수 있었다. 이 방법으로 결정한 구리와 알루미늄의 열확산도는 각각 1.20cm²/s와 0.85cm²로 이 값은 공인된 값과 일치하였다. In this work a new method determining the thermal diffusivities has been attem pted with the help of the numerical analysis of the Rosencwaig-Gersho theory. It is shown that when different backing materials are attached to a solid sample of given thicknes there exists an optical chopping frequency contributing to the same photoacoustic signal intensities regardless of the backing materials. And it is also shown that the chopping frequency depends on the sample thickness. The relation between the chopping frequency and the sample thickness shows a straight line on the log-log graph. The thermal diffusivties are determined by the distance shifted from the line of ??. The thermal diffusivities of copper and aluminium obtained by this method are 1.20cm²/s and 0.85cm²/s respectively, which agree with the values shown elsewhere.

相互作用複數目標計劃法에 關한 硏究

李昌孝 홍익대학교 경영연구소 1983 경영연구 Vol.6 No.-

대화형 ε-제약법의 적용

이창효 부산여자대학교 사회과학연구소 1998 사회과학연구 Vol.- No.2

階層化 意思決定法에 관한 硏究

李昌孝 新羅大學校 1993 論文集 Vol.36 No.1

Multicriteria decision making research in the 1970s focused on the theoretical foundations of multiple objective mathematical programming and on procedures and algorithms for solving multiple objective mathematical programming problems. During the 1980s. emphasis shifted toward the implementation of multicriteria decision making models on computers with the aid of decision support systems. However we will need simple. understandable. and usable approaches for solving multiple objective decision making and multiattribute decision making problems. As a technique not firmly rooted in utility theory. the Analytic Hierarchy Process. for the most part. has remained outside of the main stream of decision analysis research. It is based on the principle that. to make decisions. experience and knowledge of p대ple is at least as valuable as the data they use. Specially. the Analytic Hierarchy Process is a theory of measurement for dealing with quantifiable and/or intangible criteria that has found rich applications in decision theory. confilict resolution and in modeling unstructured problems in the economic. social. and administrative sciences. Decision making applications of the Analytic Hierarchy Process are carried out in two phases:hierarchic design and evaluation. As phase 1. which is perhaps th emost important aspect of the Analytic Hierarchy Process. the decision analyst should break down the decision problem into a hierarchy of interrelated decision elements. At the top of the hierarchy lies the most macro decision objective. The lower levels of the hierarchy contain attributes which contribute to the quality of the decision. The evaluation phase is based on the concept of paired comparisions. The elements in a level of the hierarchy are compared in relative terms as to their importance or contribution to a given criterion that occupies the level immediately above the elements being compared. This process of comparision yields a relative scale of measurement of the proities or weights of the elements. These relative weights sum to unity. The comparisions are performed for the elements in al level with respect to all the elements in the level above. The final or global weights of the elements at the bottom level of the hierarchy are obtained by adding all the contributions of the elements in the level above. Although the technology of large computers continues its rapid advance. it seems likely that most implementations will occur at the microcomputer level. With the recent widespread introduction of the microcomputer to the decision making process. Microcomputers have reached a stage where the Analytic Hierarchy Process can now be used effectively for solving a variety of decision problems. A few software packages. such as Criterium. and Expert Choice have become particularly popular. More recoently, application of the electronic spreadsheet for solving the Analytic Hierarchy Process problems has been presented. Clearly. the use of microcomputers for solving the Analytic Hierarchy Process problems is indeed valuable for the decision analyst who does not have easy access to a large-scale mainframe computer. To demonstrate the usefulness of the Analytic Hierarchy Process a case study of an air pollution control problem was examined. As a result. the optimal control stratigies were suggested. In the situation of existing many conflicting attributes in airpollution control. policy makers or planners may utilize the Analytic Hierarchy Process to reinforce their intuitive judgements.

수도권 창업기업의 생멸에 대한 공간분포 패턴 분석

이창효,Yi, Chang-Hyo 한국국토정보공사 공간정보연구원 2015 지적과 국토정보 Vol.45 No.1

본 연구의 목적은 창업기업의 생성과 소멸이라는 생애주기 변화와 관련하여 나타나는 공간분포 패턴의 차이를 확인하는 것이다. 이를 위하여, 본 연구에서는 전국기업체총람과 폐업기업정보를 활용하여, 2007년부터 2009년까지 수도권에서 창업한 기업의 위치정보와 생성시점, 그리고 소멸시점을 파악하였다. 창업기업의 생멸분포 패턴에 대한 분석은 공간상의 분석단위에 대한 분포패턴을 정량적으로 분석하는 방법론인 평균점, 표준거리, 타원형 표준편차, 그리고 공간자기상관도 분석기법이 적용되었다. 분석 대상 창업기업 5,810개 중, 1,322개 기업이 5년 내에 소멸함으로써 77.25%의 5년 생존율이 산정되었다. 지역별 생멸분포 패턴의 분석 결과, 수도권 전체, 서울, 그리고 인천 경기 모두에서 기업의 생성분포에 비해 소멸분포가 분산의 정도가 큰 것으로 나타났다. 또한, 기업속성별 생멸분포 패턴 분석결과, 기업의 규모와 업종에 따라 창업기업의 생성과 소멸분포에 차이가 있는 것으로 확인되었다. The purpose of this paper is to verify the spatial distribution difference between formation and dissolution of start-up firms related to life-cycle. For this purpose, a Korean business directories and information on business closures were used, and location information, starting time, and closure time of start-up firms in the Seoul metropolitan region from 2007 to 2009 were generated in this study. It applied The spatial distribution analysis methods on the formation and dissolution of the start-up firms included Barcki measurements and global spatial autocorrelation. The total number of start-up firms was 5,810, and their five-year survival rate was 77.25%. The dissolution pattern of the start-up firms was dispersed more than their formation pattern in each of the Seoul metropolitan regions, the city of Seoul, and the city of Incheon-Gyeonggi province. In addition, differences between the formation and dissolution patterns according to size and industry category of the firms were confirmed.

국내 주택시장을 고려한 도시통합모델의 주거입지 하위모델 연구

이창효,이승일 도시정책학회 2011 도시정책학회 학술대회 Vol.2011 No.3

박막형 압전진동자의 전기등가회로상수의 측정방법

李昌孝 漢陽大學校 基礎科學硏究所 1992 基礎科學論文集 Vol.11 No.-

임의로 선택한 진동수에서 압전진동자의 impedance를 LRC meter로 측정하고, 이렇게 측정한 값 중에서 임의로 선택한 4개의 data를 컴퓨터로 수치 계산하여 압전진동자의 전기등가회로상수를 결정하는 방법을 모색하였다. 본 연구에서는 rf planar magnetron sputtering장치로 c-축 방향성의 ZnO 압전박막 진동자를 제작하여 위의 방법으로 이 진동자의 전기등가상수회로와 전기역학 결합계수를 측정하였는데 그 결과 이 방법이 기존의 여러 방법보다 훨씬 편리함을 알았다. A method determining the equivalent electric circuit constants of piezoelectric vibrator was attempted by numerical calculation with arbitrary four values of the measured impedances using a LRC meter. In this study the equivalent electric circuit constants and electricmechanical coupling constant of a c-axis orientation ZnO piezoelectric thin film vibrator fabricated by rf planar magnetron sputtering equipment were measured by this method. The result showed that this method was more convenient than any conventional method.

多次元尺度構成法의 多屬性意思決定問題에의 適用

李昌孝 新羅大學校 1991 論文集 Vol.31 No.1

The problems of multiple criteria decision making can be broadly classified into two categories: multiple attribute decision making and multiple objective decision making. In actual practice this classification is well fitted to the two facets of problem solving-multiple attibute decision making is for selection, multiple objective decision making is for design. In this paper. we are focused on the applicability of multidimensional scaling tomultiple attribute decision making problems. Multidimensional scaling is a set of mathematical techniques that enable a research to uncover the hidden structure of dta bases. When alternatives have too many attributes or in some cases have vague or unknown attributes, multidimensional scaling has great advantage in the solution of multiple attribute decision making problems. The decision maker's orderings of the proximities of pairs of alternatives can be used to construct a multidimensional spatial representation. Alternatives ae represented by points in this space and the points that are close together are assumed to be close together in terms of preference. The decision maker is asked to \locate his ideal alternative in this space. and then the distance from the ideal point is measured in order to rank the alternatives in terms of preference. However this multidimensional scaling procedure suggests several points. For one thing each pair of alternatives given in the judgement must be ranked in terms of its similarity. This requires quite fine discrimination. In addition, the configuration is not unique, and the distance measure used to form the configuration assumes that the attributes are independent. In spite of the greater burden, this method is particularly useful when the number of attributes is quite large, say more than 7. and when the majority of attributes are expressed in the qualitative way.