Dissertationen zum Thema „Fuzzy multicriteria decision-making“

Continuous development has been occurring in the area of decision support systems. Modern systems focus on applying decision models that can provide intelligent support to the decision maker. These systems focus on modelling the human reasoning process in situations requiring decision. This task may be achieved by using an appropriate decision model. Multicriteria decision making (MCDM) is a common decision making approach. This research investigates and seeks a way to resolve various issues associated with the application of this model. MCDM is a formal and systematic decision making approach that evaluates a given set of alternatives against a given set of criteria. The global evaluation of alternatives is determined through the process of aggregation. It is well established that the aggregation process should consider the importance of criteria while determining the overall worth of an alternative. The importance of individual criteria and of sub-sets of the criteria affects the global evaluation. The aggregation also needs to consider the importance of the sub-set of criteria. Most decision problems involve dependent criteria and the interaction between the criteria needs to be modelled. Traditional aggregation approaches, such as weighted average, do not model the interaction between the criteria. Non-additive measures such as fuzzy measures model the interaction between the criteria. However, determination of non-additive measures in a practical application is problematic. Various approaches have been proposed to resolve the difficulty in acquisition of fuzzy measures. These approaches mainly propose use of past precedents. This research extends this notion and proposes an approach based on similarity-based reasoning. Solutions to the past problems can be used to solve the new decision problems. This is the central idea behind the proposed methodology. The methodology itself applies the theory of reasoning by analogy for solving MCDM problems. This methodology uses a repository of cases of past decision problems. This case base is used to determine the fuzzy measures for the new decision problem. This work also analyses various similarity measures. The illustration of the proposed methodology in a case-based decision support system shows that interactive models are suitable tools for determining fuzzy measures in a given decision problem. This research makes an important contribution by proposing a similarity-based approach for acquisition of fuzzy measures.

Continuous development has been occurring in the area of decision support systems. Modern systems focus on applying decision models that can provide intelligent support to the decision maker. These systems focus on modelling the human reasoning process in situations requiring decision. This task may be achieved by using an appropriate decision model. Multicriteria decision making (MCDM) is a common decision making approach. This research investigates and seeks a way to resolve various issues associated with the application of this model. MCDM is a formal and systematic decision making approach that evaluates a given set of alternatives against a given set of criteria. The global evaluation of alternatives is determined through the process of aggregation. It is well established that the aggregation process should consider the importance of criteria while determining the overall worth of an alternative. The importance of individual criteria and of sub-sets of the criteria affects the global evaluation. The aggregation also needs to consider the importance of the sub-set of criteria. Most decision problems involve dependent criteria and the interaction between the criteria needs to be modelled. Traditional aggregation approaches, such as weighted average, do not model the interaction between the criteria. Non-additive measures such as fuzzy measures model the interaction between the criteria. However, determination of non-additive measures in a practical application is problematic. Various approaches have been proposed to resolve the difficulty in acquisition of fuzzy measures. These approaches mainly propose use of past precedents. This research extends this notion and proposes an approach based on similarity-based reasoning. Solutions to the past problems can be used to solve the new decision problems. This is the central idea behind the proposed methodology. The methodology itself applies the theory of reasoning by analogy for solving MCDM problems. This methodology uses a repository of cases of past decision problems. This case base is used to determine the fuzzy measures for the new decision problem. This work also analyses various similarity measures. The illustration of the proposed methodology in a case-based decision support system shows that interactive models are suitable tools for determining fuzzy measures in a given decision problem. This research makes an important contribution by proposing a similarity-based approach for acquisition of fuzzy measures.<br>Thesis (PhD Doctorate)<br>Doctor of Philosophy (PhD)<br>School of Information and Communication Technology<br>Full Text

Este trabalho apresenta o processo de seleção de estratégia de reconversão industrial de uma pequena refinaria de petróleo no Sul do Brasil através da aplicação de uma modelagem integrada, utilizando um método multicriterial e programação matemática. Neste estudo foram identificadas linhas de negócio potencialmente aplicáveis à realidade da companhia e definido um conjunto de critérios de análise abrangendo as três dimensões da sustentabilidade empresarial: econômica, social e ambiental. Com base na avaliação da importância relativa de cada critério, atribuída por um grupo de decisores, e no desempenho das linhas de negócio em cada um dos critérios, foi aplicado método fuzzy TOPSIS para análise e ordenação das linhas de negócio. As informações resultantes desta análise, juntamente com dados econômicos objetivos, foram utilizadas em um modelo de programação linear inteira para avaliar portfólios viáveis de linhas de negócio, identificando estratégias candidatas à implementação na refinaria. O desempenho global de cada estratégia candidata, obtido mediante agregação dos desempenhos individuais das linhas de negócio e calculado conforme método fuzzy TOPSIS, foi analisado através de ferramentas gráficas, de modo a gerar elementos para subsidiar a seleção da melhor estratégia de reconversão. Os resultados obtidos demonstraram a eficiência da abordagem proposta, no sentido de facilitar o entendimento e a exploração da situação problema e, assim, oferecer um adequado suporte à tomada de decisão.<br>This dissertation presents a selection process of industrial reconversion strategy in a small oil refinery in southern Brazil by applying an integrated modeling approach, using a multicriteria and a mathematical programming method. Potentially performing business lines were identified, as well a set of criteria covering the three dimensions of corporate sustainability: economic, social and environmental. Based on the relative importance evaluation of each criteria given by a group of decision-makers, and on performance of the business lines in each of the criteria, fuzzy TOPSIS method was applied for analysis and sorting of business lines. The information resulting from this analysis, along with objective economic data, were used in integer linear programming model to evaluate effective portfolios of business lines, identifying candidate strategies to implement in the refinery. Fuzzy TOPSIS is used to generate overall performance scores of each candidate strategy, aggregating the individual performance of the business lines. The sustainability assessment was analyzed through graphical tools in order to generate information to support the selection of the best strategy for the industrial reconversion. The results demonstrated the efficiency of the proposed approach to facilitate the understanding and exploitation of the problem situation and thus offer adequate support to decision making.

A group health insurance plan is an insurance plan that provides healthcare coverage to a selected group of people. In various countries, group health insurance plans are one of the major benefits offered through employers in the private sector. In recent years, the numbers of group health insurance plans offered in the market of health insurance have been increasing rapidly. This is due to compulsory government policies, which are imposed on employers in the private sector leading to an increasing demand for this insurance plan. Accordingly, employers may face a wide variety of available group health insurance plan alternatives. Despite the fact that employers in the private sector have a crucial and significant role in the health insurance market all over the world, little is known about how employers evaluate and choose group health insurance plans to cover their employees against the payments of benefits as a result of sickness or injury. Therefore, the primary concern in this research is to propose a model to assist employers within the private sector to evaluate alternative group health insurance plans and to select the most appropriate, in order to provide the perfect health care environment for their employees. In this research, a new hybrid Fuzzy Multiple Criteria Decision Making (MCDM) model is proposed for the selection problem. The proposed model tackles some issues that may be associated with the selection of the group health insurance plan, such as modelling uncertainty, studying the dependence among decision attributes, deriving decision attributes importance weights and ranking various alternatives. In the proposed hybrid model, four extension approaches based on the Fuzzy Delphi, Fuzzy Decision Making Trial and Evaluation Laboratory (DEMATEL), Fuzzy Group Prioritisation and Fuzzy Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) methods are developed. Unlike the existing methods, the four proposed approaches, a new extended Fuzzy Delphi (FDE) method, a new extended Fuzzy DEMATEL method, a new Fuzzy Group Prioritisation (FGP) method and a new extended Fuzzy TOPSIS method, consider the importance weight of each member in group decision making since the selection problem needs evaluations from decision makers (DMs) with different levels of expertise and different perceptions. In the literature, there is some work on these methods, but to our knowledge, no research exists that combines these four methods. Moreover, the proposed model might be applied, due to its novelty, to any MCDM problem uncertainty in different. Furthermore, four new prototype decision support tools, termed Fuzzy Delphi Solver, Fuzzy DEMATEL Solver, Fuzzy Group Prioritisation Solver and Fuzzy TOPSIS Solver were developed in this study, based on the concepts of the four proposed approaches, in order to provide user-friendly interfaces for facilitating the application of these approaches. MATLAB software Version R2013a was adopted as a development environment for prototyping these new decision support tools in this study. The tools developed were validated internally by using hypothetical examples and checking the correctness of the results obtained by comparing them to other results generated from other software, such as Microsoft Excel or LINGO V13.0 software. In addition, a practical validation of the proposed hybrid Fuzzy MCDM model was investigated through conducting a case study of the Saudi health insurance industry. The main objectives of the case study were: 1) investigation of the evaluation process of selecting a group health insurance plan, including identifying the selection criteria and alternatives, studying the dependency issue, deriving the criteria weights, and ranking available alternatives; 2) application of the new decision support tools developed. In this case study, a group of nine DMs, Human Resources (HR) managers at nine different private companies in Saudi Arabia, were selected to take part of this case study. Their involvement achieved the first objective of the case study. At the end of the case study, a sensitivity analysis was conducted to indicate the robustness and the reliability of the results obtained. It is concluded that the proposed model is indeed beneficial. Finally, areas for further research were identified.

A seleção de fornecedores tem impacto significante no custo e na qualidade de produtos manufaturados. Por isso, a seleção de fornecedores passou a ser vista como uma atividade bastante crítica para o desempenho da empresa compradora. Muitos estudos da literatura propõem o uso dos métodos multicritério fuzzy TOPSIS (Technique for Order of Preference by Similarity to Ideal Solution) e fuzzy AHP (Analytic Hierarchy Process) para apoiar a seleção de fornecedores. Contudo, não são encontrados estudos que avaliem o desempenho destes métodos quando usados neste domínio de problema. Diante desta lacuna, este estudo compara os métodos fuzzy TOPSIS (CHEN, 2000) e fuzzy AHP (CHANG, 1996) no apoio à seleção de fornecedores. Esta pesquisa utiliza uma abordagem quantitativa descritiva empírica, baseada em modelagem e simulação. Os métodos fuzzy TOPSIS e fuzzy AHP foram aplicados em um caso ilustrativo de seleção de fornecedores. O desempenho dos fornecedores e o peso dos critérios foram avaliados por um especialista de uma empresa. Modelos de simulação foram implementados usando MATLAB® e aplicados na seleção de fornecedores de uma empresa de uma cadeia de suprimentos automotiva. Cinco fornecedores foram avaliados em relação à qualidade, custo, entrega, perfil e relacionamento. O peso dos critérios e o desempenho dos fornecedores foi avaliado por meio da opinião de um especialista da empresa. Posteriormente, os métodos fuzzy TOPSIS e fuzzy AHP foram comparados em relação à capacidade de apoiar a decisão em grupo, qualificação de fornecedores, escolha final de fornecedores, situações de compra e modelagem de decisões sob incerteza. A eficiência dos métodos em relação à complexidade computacional e à interação requerida com o usuário também foi comparada. Os resultados mostraram que o fuzzy TOPSIS é mais flexível e mais adequado que o fuzzy AHP para modelar diferentes tipos de cenários de seleção de fornecedores. A realização desta discussão é sugerida por Ertugrul e Karakasoglu (2008), e é relevante para ajudar pesquisadores e gestores na escolha de abordagens efetivas para lidar com diferentes cenários de seleção de fornecedores.<br>Supplier selection has a significant influence on the cost, quality and delivery of products of the buying company. Therefore, supplier selection has become a very critical activity to the performance of the buying company. Several studies presented in the literature propose the use of fuzzy TOPSIS (Technique for Order of Preference by Similarity to Ideal Solution) and fuzzy AHP (Analytic Hierarchy Process) to aid the decision process of supplier selection. However, there are no comparative studies of these two methods when applied to the problem of supplier selection. Thus, this paper presents a comparative analysis of the methods fuzzy TOPSIS (Chen, 2000) and fuzzy AHP (Chang, 1996) applied to the problem of supplier selection. A descriptive quantitative approach was adopted as the research method. Algorithms of the methods fuzzy TOPSIS and fuzzy AHP were developed in Matlab© and applied to the selection of suppliers of a company in the automotive production chain. Five suppliers were evaluated regarding quality of conformance, cost, delivery, profile and relationship. The weight of the criteria and the performance of the suppliers were evaluated by specialist opinion from the studied company. The methods Fuzzy TOPSIS e Fuzzy AHP were compared in terms of ability to support the group decision, supplier qualification, final choice of suppliers, buying situations and modeling decisions under uncertainty. The efficiency of the methods with respect to computational complexity and the required user interaction was also compared. The comparative analysis shows that Fuzzy TOPSIS presents better than Fuzzy AHP performance, especially in scenarios in wich many alternatives are evaluated. Thus, Fuzzy TOPSIS is more flexible and appropriate than Fuzzy AHP to deal with supplier selection problem. This paper presents a new study, comparing the methods Fuzzy TOPSIS and Fuzzy AHP. As commented by Ertugrul and Karakasoglu (2008), a study such as this can contribute to the advance of knowledge, helping researchers and practitioners choosing more effective approaches to supplier selection.

Building Information Modeling (BIM) is becoming imperative across building disciplines to improve communication and workflow from the first blueprint. Maintenance and facility management is however lagging behind in adoption and research of BIM. Utilizing research-through-design, this study explores BIM-enabled facility management and the critical practice of decision-making at the Celsius building in Uppsala. Contextual design and inquiry were applied to identify and suggest important design features that support decisions related to the task of establishing maximum room occupation. Results show that facility managers can make use of fuzzy multicriteria decision-making and expert heuristics to independently reach conclusions. Important design features were found to heavily rely on the existing building models, where context-view filtered to room capacity data in the existing BIM-system effectively supported the users’ assessment of data. The filtered, aggregated information presented in a simplified mobile format was insufficient for decision-making, suggesting that the building model was more important than initially perceived.

Conselho Nacional de Desenvolvimento Científico e Tecnológico<br>Sustainability is becoming a major driving force in energy policy, leading to the development of different strategies and projects. Many of these strategies are related to the application of novel methodologies for selecting Renewable Energy Systems (RES) and energy storage systems. Electrical generation with biogas from municipal solid waste (MSW) is one of the main alternatives to concern all the conceptions of sustainability - social, economic and environmental constrains. The complexity of sustainability and energy planning makes the multicriteria analysis a valuable tool for the decision making process. The use of an effective methodology for RES selection a decision making process is essential to guarantee the adequate energy management of the biogas and the MSW landfill. This methodology must be able to balance positive and negative aspects, achieving an overall solution that best satisfies the management needs. It is essential to deal with several parameters and concern the decision maker (DM) interaction over the decision making process. By applying the DM preferences into the development of the methodology, it is possible to corroborate the methodology outcome. The presented thesis will therefore develop a novel methodology for selection of RES fuelled by biogas from MSW landfills. This methodology taking as basis fuzzy multi-rules and multi-sets to provide an accurate analysis of conflicting aspects - operational, economic, environmental, social, etc. These aspects are taken into account for each study case according to different perspectives adopted by the DMs. The novel arrangements developed in this work are the creation of a previous classification of the priority criteria, the application of meta-rules and how to structure the fuzzy rules construction. The proposed arrangements have the purpose of easing the understanding of the methodology, as well as improving the DM interaction over the decision making process achieving in this way a better solution. This work presents the application of the novel decision making process to select the most appropriate energy source fuelled by biogas from MSW, considering the Caturrita II landfill located at Santa Maria City, Brazil. In conclusion, it is important to emphasize that the novel software may be used in any energy system selection, for supplying or storage, according to the analysis of several criteria and perspectives for each regional circumstances, as well as particular management needs..<br>A busca pelo desenvolvimento sustentável, em âmbitos sociais e ambientais, é um fator de extrema importância que incentiva a elaboração de várias pesquisas e projetos, como por exemplo, à aplicação de técnicas de gerenciamento e seleção de fontes alternativas renováveis de geração de energia. Dentre estas fontes, o aproveitamento energético do biogás resultante da decomposição de resíduos sólidos urbanos é um dos meios que propicia um desenvolvimento sustentável de forma mais completa. Para o devido aproveitamento de fontes alternativas renováveis, como a geração de energia elétrica e térmica a partir de resíduos sólidos urbanos, deve-se tomar como base métodos multicriteriais, considerando a existência de uma série de critérios para atender necessidades e interesses diversos quando se deseja selecionar tecnologias de geração e armazenamento de energia. A partir da utilização de métodos de ajuda a decisão é possível incorporar de forma clara as preferências dos agentes de decisão, obtendo como resposta final uma solução mais satisfatória e que pode ser corroborada através de validações heurísticas discussões dos resultados junto aos agentes de decisão. Mediante estes argumentos, o presente trabalho tem a finalidade de desenvolver uma metodologia de apoio a decisão para a seleção de sistemas para geração de energia elétrica com biogás proveniente de resíduos sólidos urbanos, avaliando devidamente cada processo decisório de acordo com aspectos econômicos, operacionais, ambientais e sociais. Para o alcance deste objetivo fez se uso da lógica difusa baseada em regras e conjuntos fuzzy aplicados sobre diversos critérios, avaliando diferentes perspectivas. Os aperfeiçoamentos mais importantes apresentados na elaboração desta tese se referem à criação de uma etapa de relevância prévia aos critérios em análise, criação e seleção de meta-regras e forma de apresentação e construção de tais regras, facilitando o entendimento dos agentes de decisão para a avaliação do processo decisório e propiciando uma maior participação dos mesmos para obtenção de um resultado mais satisfatório. É possível observar ainda que os aperfeiçoamentos desenvolvidos permitiram a devida construção e averiguação das modelagens construídas. No estudo de caso principal aterro sanitário Caturrita II localizado na cidade de Santa Maria é verificada a aplicabilidade da metodologia de ajuda a decisão desenvolvida visando a seleção da fonte de geração de energia elétrica mais apropriada a ser utilizada no aterro em questão. Por fim, deve-se enfatizar que a partir dos aperfeiçoamentos alcançados durante o desenvolvimento desta tese, foi possível construir uma metodologia de ajuda a decisão genérica que pode ser aplicada não somente na seleção de sistemas de geração de energia em aterros, mas também na seleção de quaisquer sistemas de geração e armazenamento de energia, desde que todos os aspectos envolvidos no processo decisório sejam devidamente incorporados no problema em questão.

碩士<br>義守大學<br>資訊管理學系碩士班<br>95<br>Through the significant improvement of technology, the information technology keeps weeding through the old to bring fort the new. In order to go with the stream, increase operation efficiency, maintain the more complex information system availability, and update functions frequently, most corporations need the assistance from professional companies. To outsource the information system which is not the core ability of corporations, not only can reduce the costs of building system, but help corporations to focus on their particular field. Besides, choosing an appropriate contractor will raise the successful rate of outsourcing. Owing to the qualitative data and quantization data coexistent is unavoidable, and most of them are fuzziness and uncertainty. Therefore, this study planed to combine the Fuzzy Logic and the Fuzzy Multicriteria Decision-Making methods, which applies the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) with fuzzy number and Fuzzy Preference Ranking Organization METHod for Enrichment Evaluation (Fuzzy PROMETHEE) to have the best result of contractor evaluation. The case of selection for information system outsourcing in this study was evaluated by four decision-makers, and based on Linguistic Variable to give the weight of seven criteria which include the experience of contractor, the goodwill of contractor, the cooperative ability, technology skills, sustainable ability, and management ability of contractor, and general problems. Furthermore, according to criteria, the four candidature outsourcing contractors would be evaluated by linguistic variables. Additionally, depends on the Fuzzy Set Theory to carry on the set operation and de-fuzzify. Finally, obtains the rank of candidature outsourcing contractors. The application of FMCDM to evaluate the information system outsourcing contractor can be consequent on a common consensus at group decision making by rationalization and systematization, as efficient solve the lack of flexibility in the traditional FMCDM. The result of rank will provide the reference to the corporation which may need assistance of information system outsourcing to improve the efficiency.

碩士<br>朝陽大學<br>資訊管理系碩士班<br>87<br>Because the traditional multiple attribute decision methods have to request the crisp number from the decision maker which number is difficult to express the linguistic characteristic of the criteria or the weights. Therefore, many studies utilize fuzzy set theory in the application of decision making to resolve this problem. The decision method to handle fuzzy (non-crisp) criteria is hence called Fuzzy Multiple Attribute Decision Making (FMADM). Generally, the process of FMADM consists of two important parts : the fuzzy aggregation judgment and the fuzzy ranking method. In accordance with the FMADM, our research intends to solve the following problems. 1. Traditional fuzzy multi-attribute decision model is incapable of representing the default logic, i.e., the decision logic using informal criterion. In general, it is the major criteria determine the process result, however, second-order (minor) criteria after the process of the major criteria should also be considered and integrated into the model. 2. It is difficult to assign the proper weights for the informal criteria. 3. Each ranking method has its advantage and disadvantage and there is no general model to deal with various type of problems concerning with fuzzy ranking. Therefore, an intelligent system to help ranking fuzzy numbers is needed. We will integrate Yager''s method with second-order structure concept to handle the prioritized criteria. A fuzzy ranking rule base will be constructed to solve the ranking problem. Finally, we will implement an intelligent fuzzy system to support the traditional fuzzy multi-attribute decision model.

碩士<br>國立臺灣科技大學<br>資訊工程系<br>95<br>In this thesis, we present a new approach for fuzzy risk analysis based on the ranking of fuzzy numbers. First, we present a new method for ranking fuzzy numbers using the α-cuts, the belief features and the signal/noise ratios of fuzzy numbers. The proposed method can overcome the drawbacks of some existing methods for ranking fuzzy numbers. Then, we apply the proposed fuzzy ranking method to present a fuzzy risk analysis algorithm to deal with fuzzy risk analysis problems. The proposed fuzzy risk analysis algorithm can provide a useful way to deal with fuzzy risk analysis problems. In this thesis, we also present a new method for prioritized multicriteria decision making, where the weights of the lower priority criteria of each alternative depends on whether each alternative satisfies the requirements of all the higher priority criteria or not. If the requirements of all the higher priority criteria can not be satisfied by the alternative, then the weights of the lower priority criteria are all zero. That is, the degrees of satisfaction with respect to the lower priority criteria do not affect the overall degree of satisfaction. Furthermore, we present a generalized prioritized multicriteria decision making method for handling multicriteria decision making problems in which some criteria may have equal priority and the criteria with equal priority are aggregated by using the ordered weighted averaging (OWA) operator or the weighted averaging method. The proposed methods can overcome the drawbacks of the Yager’s methods. The proposed generalized multicriteria decision making method can handle multicriteria decision making problems in a more intelligent and more flexible manner.

博士<br>國立臺灣大學<br>環境工程學研究所<br>94<br>The uncertainty of the AHP methodology is analyzed in a systematic view in this study. This work addresses two kinds of uncertainties associated with the AHP: the first is uncertainty associated with normalization; the other is the uncertainty of evaluation of qualitative criteria. The effect of fuzzy linguistic variables methods are also examined in this investigation. Simulation experiments are developed to quantify the uncertainty of the AHP. Computational results reveal that (1) the rank reversal phenomenon can occur no matter whether an alternative is added or not; (2) the procedure of normalization is the principal source of the uncertainty of the AHP methodology; (3) the fuzzy linguistic variables methods can reduce the uncertainty of the AHP. The transmission of uncertainty of the AHP is also described in this study. Environmental management problems are very complex and require considering numerous factors, such as environmental, economic, and social aspects. Qualitative and quantitative data always exist simultaneously in real world decision making situations. A novel multiobjective programming approach is proposed in this study to solve qualitative and quantitative objectives for environmental management problems. This approach integrates the multiattribute and multiobjective decision making methods and contains three main steps to solve the multiobjective programming problems, including formulation of the decision model, the alternatives prioritization by the fuzzy AHP method, and solving the model. This study also reviews several models developed to support decision making in municipal solid waste management (MSWM). The concepts underlying sustainable MSWM model can be divided into two categories: one incorporates social factors into decision making methods, and the other includes public participation in the decision-making process. The public is only apprised or takes part in discussion and has little effect on decision making in most researches. Few studies have considered public participation in the decision-making process. Additionally, all the methods seek to strike a compromise between concerned criteria, not between stakeholders. However, the source of the conflict arises from the stakeholders’ complex web of value. Such conflict affects the feasibility of implementing any decision. The purpose of this study is to develop a sustainable decision making model for MSWM to overcome these shortcomings. The proposed model combines multicriteria decision making (MCDM) and a consensus analysis model (CAM). The CAM is built up to aid decision-making when MCDM methods are utilized and, subsequently, a novel sustainable decision making model for MSWM is developed. The main feature of CAM is the assessment of the degree of consensus between stakeholders for particular alternatives. Two case studies for food waste management and fly ash of municipal solid waste incinerator in Taipei are presented to demonstrate the practicality of this model.

博士<br>國立臺灣科技大學<br>資訊工程系<br>93<br>Fusion and aggregation of information are important topics in many researches, such as fuzzy logic systems, multi-attribute decision-making, group decision-making, and information retrieval,…, etc. In this dissertation, we firstly present a new similarity measure of generalized fuzzy numbers. First, we present a method called the Simple Center of Gravity Method (SCGM) to calculate the center-of-gravity (COG) points of generalized fuzzy numbers. Then, we use the SCGM to propose a new method to measure the degree of similarity between generalized fuzzy numbers. The proposed similarity measure uses the SCGM to calculate the COG points of trapezoidal or triangular generalized fuzzy numbers and then to calculate the degree of similarity between generalized fuzzy numbers. We also prove some properties of the proposed similarity measure and use an example to compare the proposed method with the existing similarity measures. The proposed similarity measure can overcome the drawbacks of the existing methods. We also apply the proposed similarity measure to develop a new method to deal with fuzzy risk analysis problems. The proposed fuzzy risk analysis method is more flexible and more intelligent then the existing methods due to the fact that it considers the degrees of confidence of decision-makers’ opinions. Furthermore, we modify the proposed similarity measure to simplify the calculation process to measure the degree of similarity between generalized fuzzy numbers. We also present a method to measure the degree of similarity between interval-valued fuzzy numbers. In this dissertation, we also present a new method for ranking generalized fuzzy numbers. Based on the proposed method, we also present an algorithm to deal with fuzzy risk analysis problems. The proposed method considers the centroid points and the standard deviations of generalized fuzzy numbers for ranking generalized fuzzy numbers. We also use an example to compare the proposed method with the existing centroid-index ranking methods. The proposed ranking method can overcome the drawbacks of the existing centroid-index ranking methods. The proposed fuzzy risk analysis algorithm can overcome the drawbacks of the one we presented in the above. In this dissertation, we also use fuzzy numbers to extend the traditional Induced OWA (IOWA) operator to present the fuzzy-number IOWA (FN-IOWA) operator, where fuzzy numbers are used to describe the argument values and the weights of the FN-IOWA operator, and the aggregation results are obtained by using fuzzy number arithmetic operations. Based on the proposed FN-IOWA operator and the proposed ranking method of fuzzy numbers, we present a new algorithm to deal with multi-criteria fuzzy decision-making problems. The proposed algorithm can deal with multi-criteria fuzzy decision-making problems in a more intelligent and more flexible manner. Furthermore, we use the FN-IOWA operator and linguistic quantifiers to present a new information fusion algorithm for fusing fuzzy opinions in heterogeneous group decision-making environment. The proposed information fusion algorithm has the following advantages: (1) It uses linguistic quantifiers based on the FN-IOWA operator to flexibly determine the weight wi of the opinion of each expert Ei for aggregating the experts’ fuzzy opinions. (2) The experts’ opinions do not necessarily need to have a common intersection. (3) It does not need to use the Delphi method to adjust fuzzy numbers given by experts. In this dissertation, we extend the prioritized operator presented by Yager and to present a prioritized information fusion algorithm based on the similarity measure of generalized fuzzy numbers. The proposed prioritized information fusion algorithm has the following advantages: (1) It can handle prioritized multi-criteria fuzzy decision-making problems in a more flexible manner due to the fact that it allows the evaluating values of criteria to be represented by generalized fuzzy numbers or crisp values between zero and one, and (2) it can deal with prioritized information filtering problems based on generalized fuzzy numbers. Furthermore, we present a new prioritized information fusion algorithm for handling information filtering problems based on interval-valued fuzzy numbers. Furthermore, we use the proposed fusion algorithm for handling multi-level information filtering problems. The proposed prioritized information fusion algorithm can deal with information filtering problems in a more flexible manner due to the fact that it not only can deal with information filtering problems based on interval-valued fuzzy numbers, but also can deal with multi-level information filtering problems. Finally, we point out that there are some drawbacks in the existing averaging operators (i.e., P-Norm operators, Infinite-One operators, and Waller-Kraft operators) to deal with AND and OR operations of fuzzy information retrieval. Furthermore, we present new averaging operators based on geometric-mean averaging (GMA) operators to deal with these drawbacks. We use some examples to compare the proposed GMA operators with the existing averaging operators. We also prove some properties of the proposed GMA operators. The proposed GMA operators can overcome the drawbacks of the existing averaging operators and easily determine an appropriate value of the parameter α, where α is either 0 or 1, for handling AND and OR operations of fuzzy information retrieval. Furthermore, we present generalized fuzzy number geometric-mean averaging operators (GFNGMA operators) for dealing with queries based on generalized fuzzy numbers. Furthermore, we use GFNGMA operators to deal with queries represented by interval-valued fuzzy numbers. The proposed GFNGMA operators can deal with fuzzy-number queries in a more flexible and more intelligent manner.

碩士<br>義守大學<br>資訊管理學系碩士班<br>94<br>21 centuries is a new century that enterprises are facing a swift speed of changes. The Internet revolution has triggered the competition of utilizing time and space, that forcing enterprises have no choice but to squarely face such issues of global competition, division of labor among nations and the multinational operation. Thoroughly grasping information both inside and outside the organization and quickly adjusting the intension of organization may become the key point for breaking through siege to survive. In order to fulfill such needs, the implementation of ERP systems can be the cornerstone for enterprise’s sustainable development. Although many organizations have implementing various ERP systems, greatly parts of enterprises still be learning how to fully develop the capability of ERP systems, or even face embarrassing situation of failing the implementation. To probe these problems, many reasons related to the complexity of ERP systems, enterprise scales, business process, organization cultures and consulting firms. This research concentrates on the construction of the best implementing model of ERP system, reducing various risks and obstacles within the implementing process period, making enterprise take advantages of ERP systems. After collecting the ERP experts experience and standpoints by means of the questionnaire, this research applies the fuzzy multicriteria decision-making(FMCDM) to evaluate ERP systems development methods and implementation strategies, and further constructs the best model for enterprises to implement ERP systems in order to provide enterprises a reference resource for future implementing. In addition, because of realizing in the real decision environment that qualitative and quantitative data could not coexist and full of fuzziness and uncertainty, this research integrates fuzzy logic and multicriteria decision-making to develop Fuzzy VIKOR methodology and Fuzzy PROMETHEE methodology. Those methodologies can properly mediate the conflicts and contradictions during the decision-making process, effectively act in response to the lack of flexibility while adopting traditional multicriteria decision-making to deal with fuzzy problems and benefits in multicriteria group decision-making analysis for extensive application.

碩士<br>國立成功大學<br>工業與資訊管理學系碩博士班<br>100<br>Project management is different from most aspects of business management and has attracted more and more attentions in firms. Moreover, there are currently many related software applications that can solve problems like scheduling and controlling. However, project management not only produces data, but also measures and analyzes the performance and the result of project execution. Based on obtained information and guidelines, decision makers can then plan projects more completely in the future. Conflicts between project goals often arise during project execution and how to trade off between goals in order to overcome such conflicts is often the most difficult task that project managers and project groups face. As a result, how to develop an effective method to measure project performance to aid in trade-off decisions is important issue in project management. In this study, we use fuzzy multicriteria decision making approaches to measure the performance of project management. First, we use the Fuzzy Analytic Network Process (FANP) which contains the concept of ex ante weight to develop the index and weights of dependent criteria, based on the subjective views of senior managers. We also use the concept of ex post weight to calculate the performance efficiency of projects using fuzzy data envelopment analysis. Finally, we analyze the results from these two different methods and develop the project performance assessment process. With the help of this feedback process, senior managers can better assign projects and achieve better results.

碩士<br>國立臺灣科技大學<br>資訊工程系<br>99<br>In recent years, some researchers proposed fuzzy multiple attributes group decision making methods based on ranking interval type-2 fuzzy sets. In this thesis, we present a new method for fuzzy multiple attributes group decision making based on ranking interval type-2 fuzzy sets. First, we present a new method for ranking interval type-2 fuzzy sets. Then, we present a new method for multiple attributes group decision making based on the proposed ranking method of interval type-2 fuzzy sets. In this thesis, we also present a new method for multicriteria fuzzy decision making based on ranking interval-valued intuitionistic fuzzy sets, where interval-valued intuitionistic fuzzy values are used to represent evaluating values of the decision-maker with respect to alternatives. First, we propose a new method for ranking interval-valued intuitionistic fuzzy values. Based on the proposed fuzzy ranking method of interval-valued intuitionistic fuzzy values, we propose a new method for multicriteria fuzzy decision making. The methods presented in this thesis provide us useful ways for dealing with fuzzy multiple attributes group decision making problems and multicriteria fuzzy decision making problems.

碩士<br>國立臺灣科技大學<br>資訊工程系<br>103<br>Multicriteria decision making based on intuitionistic fuzzy sets is an important research topic. In recent years, some methods have been presented for multicriteria decision making based on intuitionistic fuzzy sets. In this thesis, we propose a new similarity measure between intuitionistic fuzzy sets based on the centroid points of the transformed right-angled triangular fuzzy numbers and prove some properties of the proposed similarity measure between intuitionistic fuzzy values. Then, we propose a new multicriteria decision making method based on the TOPSIS method and the proposed similarity measure between intuitionistic fuzzy sets to overcome the drawbacks of the existing methods, where the existing methods have the drawbacks that they cannot get the preference order of alternatives in some situations due to the fact that they have “the division by zero problem”. We also use some examples to illustrate the proposed multicriteria decision making method can overcome the drawbacks of the existing methods. The proposed multicriteria decision making method provides us with a useful way for multicriteria decision making in intuitionistic fuzzy environments.

碩士<br>國立臺灣科技大學<br>資訊工程系<br>101<br>In this thesis, we present a new method for multicriteria linguistic decision making based on the aggregation of hesitant fuzzy linguistic term sets and the α-cuts of fuzzy sets. A context-free grammar is used by the expert to produce linguistic assessments of alternatives with respect to criteria. The linguistic expressions are transformed into hesitant fuzzy linguistic term sets by a transformation function. First, the fuzzy sets in each hesitant fuzzy linguistic term set are combined into a fuzzy set. Then, the system performs the α-cut operations to these aggregated fuzzy sets to get intervals, respectively, where α∈(0, 1]. Then, for each alternative, the system performs the minimum operations among the intervals obtained by the α-cuts of the aggregated fuzzy sets to get a derived interval of each alternative, where α∈(0, 1]. Finally, for each alternative, the system calculates the likelihood p(X ≥ [0, 1]) of X≥[0,1], where X is a derived interval of the alternative. The larger the likelihood value of the alternative, the better the preference order of the alternative. The proposed method is simpler and more flexible than the existing methods for multicriteria linguistic decision making based on hesitant fuzzy linguistic term sets.

碩士<br>國立交通大學<br>資訊管理研究所<br>81<br>The primary objective of this study is to integrate data base system , expert system and fuzzy multicriteria decision making method to implement a master production scheduling techniques decision system(MPSTDS) for assemble-to-order(ATO) products. Automobiles and personal computers are examples of ATO products . The ATO company sells a wide variety of customized products requiring some customer interface regarding product specifications . Competition may exist among the producers of similar products such that the firms'' sales volumes may be seriously affected by variations in quality, price,lead tile..etc. This paper proposes the MPSTDS to support the manufacture manager who can dynamically select the best suitable technique for designing a master production scheduling that according to current marketing demand , product structure type and company manufacturing stategy. Multicriteria decision making bas been one of rapid growing research areas in evaluation method . This method cooperate with fuzzy inference used to determine the best suitable technique for ATO firms. Additionally, this study uses the expert system development shell - LEVEL 5 OBJECT , dBASE III Plus and CLIPPER lanauage to develop a MPSTDS prototype system and to verify the MPSTDS. Keyword : Master Production Scheduling , Fuzzy inference , Multicriteria decision making