Instance-based learning methods like the nearest neighbour classifier have been proven to perform well in pattern classification on many fields. Despite their high classification accuracy, they suffer from high storage requirement, computational cost and sensitivity to noise. In this paper, we present a data reduction method for classification techniques based on entropy-based partitioning and center instances. Experimental results show that the new algorithms achieve a high data reduction rate as well as classification accuracy.
It is very important for a country in view of establishing national competition to make strategies for expanding technology development and generating industry technology manpower through cooperative projects between industry and universities based on their confidence and agreements. The preparation for industry technical experts with creativity and specialty is the energy necessary at this moment towards Gross National Product 20,000 dollars. This research considers diverse cooperative projects between industry and universities and successful strategies.
Depending on the nature of a product, a design process for an ergonomic product follows a series of simple or complex stages. This paper suggested a rather practical and simple design process consisting of eight stages, i.e. design needs assessment, ergonomic guidelines, anthropometry, brainstorming and idea sketch, preliminary model, drafting and rendering, working prototype, and user trials. The feasibility of this design process was verified by applying it to the development of a modified clamping hand tool handle. The case study showed how ergonomics played a role and how ergonomists overcame design difficulties by integrating good ergonomic design practices.
Zohar(1980) began to emphasize the influence of social, organizational or psychological context in occupational safety and health study. With this research trend, Neal, Griffin, and Hart(2000) developed sequential model of safety climate. This paper tried to review their model of relationships among safety climates, safety knowledge, safety motivation, and safety performance(safety compliance and safety participation).
Based on the responses from 207 manufacturing workers, this paper reviewed the proposed relationships. The results of statistical analysis showed that leadership has affirmative effects on safety knowledge and motivation, and precaution activities only on safety knowledge, safety system only on safety motivation. And also, safety knowledge and motivation have affirmative effect on safety compliance and participation.
The main purpose of this study is to identify the factors that affect purchase intention in B2C cooperative buying. The purchase intention is described in terms of participation frequency and total sum of purchasing on cooperative buying.
It is very interesting that the influencing factors are not identical when two dependent variables. participation frequency and total sum of purchasing, are used. Residential area, sex, and product variability were selected as key influencing factors for the former case. On the other hand, residential area, price, and product quality were selected as key influencing factors for the latter case. Finally, it is worth while to stress that online community, one of the factors considered, will be a key factor as market of cooperative buying is getting bigger.
The life change unit(LCU) models by behavior scientific approach method are very important in order to minimize industrial accident by the job stress. Especially, these models are analyzed the accident proneness concept through the LCU factors. Therefore this paper describes the analytical results between the Eastern LCU model and Western LCU model and its application.
Death of parents, death of close friend, jail term, personal injury or illness between the Eastern LCU models and Western LCU models are different in the human behavior and psychological sides, whereas changes in family member s health are turned out same life ideas. In the case of Korean LCU models between the unmarried and married workers living in the middle area are approved in the order of death of parents, death of spouse, death of close family member, and death of close friend.
This study presents the new dispatching rules of job shop scheduling with unbalanced machine workloads to decrease mean flow time and mean tardiness. The proposed dispatching rules consider the information related to work remaining, modified job due dates, modified operation due dates and machine workload. The performance of the new dispatching rules is compared and analyzed with the existing rules through the computer simulation at different levels of workload imbalanced. The results can be useful to the researchers and practitioners of job shop scheduling with unbalanced machine workloads.
신제품의 개발에 있어서 고객이 기대하는 품질과 신뢰성을 보증하는 것은 특별히 중요한 사항이다. 제품의 신뢰성과 수명은 개발에서 최우선으로 고려해야 할 목표이며, 이를 위해 신뢰성공학의 응용기술은 신제품개발에 있어서 매우 중요한 역할을 한다.
본 연구의 목적은 신뢰성이론을 적용하여 제품의 개발단계에 있어서 신뢰성목표 설정에서 부터 예측, 검증, 그리고 분석에 이르는 실제적인 응용에 대한 총괄적인 사항들을 보여주는데 있다. 본 논문에서는 신뢰성과 유효수명에 대한 사양결정과 예측 및 분석에 응용할 수 있는 방법들이 설명되어 있으며, 이는 제품과 부품을 개발하기 위한 신뢰성프로그램을 설립하고 시행하기 위해 가장 적합한 접근방법을 보여주고 있다.
대학과 기업이 신뢰와 협력을 기반으로 산학협력 사업을 통한 기술개발 확대와 산업기술인력 양성을 위한 전략은 국가 경쟁력 확보차원에서 매우 중요하다. 국민소득 2만불로 향하고 있는 시점에 필요한 소프트한 동력은 전문적이며 창의적인 산업기술인력 확보다. 본 연구는 정부는 다양한 산학협력사업과 산업기술인력 정책을 살펴보고 대학과 기업의 역할을 제시하고자 한다.
This paper uses a data mining methodologies to improve and predict cause of defect process variables in manufacturing process. Traditional statistical process control (SPC) techniques of control charting are not applicable in many process industries because it is difficult to analyze the cause of many process variables. The paper suggests that data mining methodologies useful when sequence rule, SVM (classification) methods are find out cause of defect process variables and SVM (prediction) methods used to predict process variables in manufacturing process. Therefore, it is allowing improved control in manufacturing process.
This study developed the planning process for technical management of engineering project applying systems engineering process in order to correctly design technical management activities early during the system development projects. The developed process describes a systemic design process of a product to be developed as well as the development process and the personnel team of project organization. As an outcome of the process implemented by a Model-Based Systems Engineering(MBSE) software, a systems engineering management plan(SEMP) and a risk management plan can easily be produced.
A centralized inventory system provides a number of stores with cost reduction, information sharing. To build centralized warehouses, two things should be considered: how many warehouses are required and where these are located. The number and location of centralized warehouses can be determined in minimizing total costs and then these cost reductions should be allocated fairly to each store. Since it is too time-intensive to find an optimal solution of minimizing total costs, this paper presents heuristics and compares its results with an optimal solution.
최근 몇 년간 인터넷 쇼핑몰은 급격한 성장세를 보이며 안정적 유통채널로 자리잡게 되었다. 인터넷 쇼핑몰의 규모가 커지면서 고객의 수는 기하급수적으로 증가하게 되었고 수천 또는 수만의 고객을 보유하는 것이 일반적인 현상이 되었다. 그러나, 인터넷 쇼핑몰의 효율적인 운영과 수익성 확보를 위해서는 현재 고객의 구성이 적절한지 전략적 측면에서 접근하여 평가하는 것이 요구된다. 즉, 기업이 현재 수행하고 있는 고객 전략이 적절한지 혹은 부적절한지에 관한 평가는 매우 중요하다고 할 수 있다. 이러한 사업적 중요성에도 불구하고 인터넷 쇼핑몰에 있어서 고객 구성 평가가 제대로 이루어지지 않고 있으며, 관련 연구도 미흡한 실정이다.
본 연구는 인터넷 쇼핑몰의 고객 구성이 적절한지 평가하도록 지원하는 고객구성평가 프레임웍을 제안한다. 제안하는 고객구성평가 프레임웍은 데이터베이스에 있는 주문 데이터와 고객 데이터 등을 기반으로 분석용 데이터를 구성한 후, 고객유지성, 고객수익성, 고객확보성, 고객전환성의 4가지 요인을 설정하고, 이하 22개의 측정 지표들을 도출함으로써, 이들을 종합한 고객구성평가 스코어카드를 구성하였다. 본 연구는 제안한 프레임웍을 국내 대표적인 커뮤니티 기반쇼핑몰에 적용하였으며, 평가 결과는 해당 쇼핑몰의 효율적인 운영을 위한 중요한 전략정보로 활용될 것으로 기대된다.
In this paper we extend the classical decision model under uncertainty to a more general case. We propose an expected utility-uncertainty model and we can make a decision by trading off between a measure of uncertainty and a measure of expected value. As a risk analysis model, the expected utility-uncertainty model can be seen to be reasonable and flexible for states of nature or individuals' preferences. Moreover, the model can explain some decision paradoxes.
This paper is intended to develop a Bayesian decision model for the repair of deteriorating system. A non-homogeneous Poisson process with a power law failure intensity function is used to describe the behavior of the deteriorating repairable system. The decision on whether to have minimal repair or imperfect repair should be made on the occurrence of a failure. However, it is difficult to make a reasonable decision due to many uncertainties intrinsic in repair actions. In this paper, prior distributions are used in order to analyze the uncertainties embedded in the decision alternatives. Especially, a prior distribution for imperfect repair with probabilistic reduction in the failure intensity is proposed. In addition, mathematical expressions to calculate the expected prior loss of each repair alternative are proposed.
최적화 문제는 일반적으로 복수개의 목적식을 가지며, 이러한 목적식들의 대부분은 서로 충돌한다. 즉, 한 개의 목적식을 최적화하면 다른 목적식들은 최적화되지 못한다. 그러므로 하나의 목적식을 최적화하는 결정변수들이 다른 목적식들을 동시에 최적화시키기가 매우 어렵다. 따라서 최적화 개념도 하나의 목적식을 고려하는 경우와는 다른 관점에서 고려해야 한다.
본 연구에서는 다목표 최적화 문제를 해결하기 위한 새로운 실수코딩 유전자 알고리즘을 제시하고, 알고리즘의 효율 평가를 위해서 다목표 유전자 알고리즘에서 가장 일반적으로 사용되고 있는 MOGA 기법과 비교한다. 제시되는 다목표 실수코딩 유전자 알고리즘에서는 여러 개의 목적식을 평가하기 위한 적합도 함수를 제안하며, 목적식들을 만족하는 다양한 파레토 최적 집합을 구축하기 위한 방안을 제시한다. 개발된 다목표 최적화 알고리즘과 MOGA 기법의 효율 평가를 위해 두 알고리즘이 파레토 최적해의 집합을 어떻게 구성하는지 비교한다. 실수코딩 유전자 알고리즘의 실험을 위해 교배연산자는 단순교배 기법을 사용하고 돌연변이 연산자는 균등돌연변이 기법을 사용한다.
우리나라 대부분의 국가산업단지 개발은 기업의 수요에 의한 개발이 아닌, 정부차원에서 공급자 위주로 이루어져왔다. 기업의 다양한 수요를 수렴하지 못한 개발은 지금의 산업단지 수요ㆍ공급의 심각한 불균형 현상을 초래하게 되었고, 특히 북평 산업단지는 95년에 준공되었지만 전국 42개 국가산업단지 가운데 현재까지 미분양률이 약 63%로서 가장 높은 상황이다.
이에 본 연구에서는 북평 산업단지 활성화를 위한 방안으로서 클러스터 구축 및 공동물류 시스템 운영을 통한 활성화 방안에 대하여 연구하였다.
In recent years, advancement of science technology and ubiquitous use of networking have led to a great increase of the need for high speed wireless internet. This has encouraged the existing Wireless Local Area Network(WLAN) system to build a new territory in the information and communication market. It provides with reasonable price and convenience, however, its connectivity, transmission speed and mobility do not meed consumers' expectations. In this paper, the standards for grading quality according to the purposes of the WLAN service, applying the concept of composition cause of the service quality that has already been agreed upon, will be reconstructed. Also, this study will try to measure the achievements of the actual WLAN service quality by utilizing the realized measurement standards.
Machining processes generate severe heat, which deforms thin metallic parts while they are being processed. The amount of deformation of a thin part is mainly dependent on heat and the shape of a component. The selection of machining parameters can decrease the temperature arisen during a machining process. This paper investigates selecting machining parameters to minimize the heat-deformation that is produced during machining. Taguchi method is applied to setup experiments. Three factors and three levels for each factor in the experimental design are selected. This research analyzes the effects of machining parameters to minimize the heat-deformation through examining the analysis of variance (ANOVA) and the signal to noise ratio (S/N).