Firms or organisations implement performance assessment to improve productivity but evaluating the performance of firms or organisations may be complex and complicated due to the existence of conflicting objectives. Data Envelopment Analysis (DEA) is a non-parametric approach utilized to evaluate the relative efficiencies of decision making units (DMUs) within firms or organizations that perform similar tasks. Although DEA measures the relative efficiency of a set of DMUs the efficiency scores generated do not consider the decision maker’s (DM’s) or expert preferences. DEA is used to measure efficiency and can be extended to include DM’s and expert preferences by incorporating value judgements. Value judgements can be implemented by two techniques: weight restrictions or constructing an equivalence Multiple Objective Linear Programming (MOLP) model. Weight restrictions require prior knowledge to be provided by the DM and moreover the DM cannot interfere during the assessment analysis. On the other hand, the second approach enables the DM to interfere during performance assessment without prior knowledge whilst providing alternative objectives that allow the DM to reach the most preferred decision subject to available resources. The main focus of this research was to establish interactive frameworks to allow the DM to set targets, according to his preferences, and to test alternatives that can realistically be measured through an interactive procedure. These frameworks are based on building an equivalence model between extended DEA and MOLP minimax formulation incorporating an interactive procedure. In this study two frameworks were established. The first is based on an equivalence model between DEA trade-off approach and MOLP minimax formulation which allows for incorporating DM’s and expert preferences. The second is based on an equivalence model between DEA bounded model and MOLP minimax formulation. This allows for integrating DM’s preferences through interactive steps to measure the whole efficiency score (i.e. best and worst efficiency) of individual DMU. In both approaches a gradient projection interactive approach is implemented to estimate, regionally, the most preferred solution along the efficient frontier. The second framework was further extended by including ranking based on the geometric average. All the frameworks developed and presented were tested through implementation on two real case studies.

In this work, it is presented the multiple objective networkflow problems. This kind of problem is converted into singleo bjective problem and solved by using the primal dual interior point method. The linear system associated to the interior point method is solved by using the Cholesky decomposition, implemented in MATLAB code. Networks of different dimensions are constructed and the computational results show the efficiency of the mentioned interior point method for solving multiple objective network flow problems.

This dissertation examines the identity threats perceived by individuals and organizations in Online Social Networks (OSNs). The research constitutes two major studies. Using the concepts of Value Focused Thinking and the related methodology of Multiple Objectives Decision Analysis, the first research study develops the qualitative and quantitative value models to explain the social identity threats perceived by individuals in Online Social Networks. The qualitative value model defines value hierarchy i.e. the fundamental objectives to prevent social identity threats and taxonomy of user responses, referred to as Social Identity Protection Responses (SIPR), to avert the social identity threats. The quantitative value model describes the utility of the current social networking sites and SIPR to achieve the fundamental objectives for averting social identity threats in OSNs. The second research study examines the threats to the external identity of organizations i.e. Information Security Reputation (ISR) in the aftermath of a data breach. The threat analysis is undertaken by examining the discourses related to the data breach at Home Depot and JPMorgan Chase in the popular microblogging website, Twitter, to identify: 1) the dimensions of information security discussed in the Twitter postings; 2) the attribution of data breach responsibility and the related sentiments expressed in the Twitter postings; and 3) the subsequent diffusion of the tweets that threaten organizational reputation.

[pt] Esta dissertação investiga um sistema de apoio à decisão baseado em Algoritmos Genéticos e Simulação Monte Carlo para a formação de carteiras de projetos de petróleo e gás. O objetivo do trabalho é avaliar o desempenho de Algoritmos Genéticos -AG- para selecionar projetos que formarão a carteira. A construção de carteiras de projetos é um problema de múltiplos objetivos, onde se deseja escolher um conjunto de projetos com perspectivas de lucro para formar uma carteira. O sistema emprega o Algoritmo Genético para formação de carteiras de projetos. Em seguida, a Simulação de Monte Carlo é utilizada para obter a função de distribuição do Valor Presente Líquido -VPL- da carteira baseado nas distribuições dos projetos escolhidos. Por último, avalia-se a carteira usando-se o método de minimização de energia que busca o equilibro dos três objetivos considerados. O problema consiste, basicamente, em maximizar a média do VPL, que representa o retorno esperado, minimizando-se o Desvio Padrão, que é a medida de risco, e maximizando-se o Percentil 90 -P90-, que significa a possibilidade de obter um maior lucro. Nos estudos de casos são apresentados os resultados da aplicação do sistema para diferentes grupos de projetos, constituídos por 16, 18, 20 e 26 projetos, onde cada um deles tem distribuições teóricas do VPL definidas por funções: F, Normal e Logarítmica, formadas por 500 dados. Os resultados obtidos mostram a eficiência do AG com a técnica de múltiplos objetivos, na utilização para a otimização de carteiras de projetos de investimento em petróleo e gás. / [en] This thesis investigates a system of support to the decision based on Genetic Algorithms and Monte Carlo Simulation for the creation of portfolio projects of oil and gas. The objective of this work is to evaluate the performance of Genetic Algorithms -GA- to select projects that will form the portfolio. The portfolio construction of projects is a problem of objective multiples, where it is wishes to choose a set of projects with profit perspectives to form a portfolio. The system uses the Genetic Algorithm to create the portfolio formation of projects. After that, the Monte Carlo Simulation is used to get the function of distribution of the Net Present Value -NPV- of the portfolio based on the distributions of the chosen projects. Finally, the portfolio is evaluated portfolio by using itself the method of minimizes energy for the three considered objectives. The problem consists, basically, in maximizing the average of the NPV which represents the return expected, minimizing the Standard of Deviation, which is the measure of the risk, and maximizing the Percentile 90 -P90-, which means the possibility to get a bigger profit. In the study of cases, it is presented the results of the application of the system for different groups of projects, consisting in 16, 18, 20 and 26 projects, where each project has theoretical distributions of the NPV defined by functions: F, Normal and Logarithmic, formed for 500 data. The gotten results show the efficiency of the GA with the technique of objective multiples, in the use of the optimization of the portfolio projects oil and gas investment.

Construction projects are complex endeavors that require the involvement of different professional disciplines in order to meet various project objectives that are often conflicting. The level of complexity and the multi-objective nature of construction projects lend themselves to collaborative design and construction such as integrated project delivery (IPD), in which relevant disciplines work together during project conception, design and construction. Traditionally, the main objectives of construction projects have been to build in the least amount of time with the lowest cost possible, thus the inherent and well-established relationship between cost and time has been the focus of many studies. The importance of being able to effectively model relationships among multiple objectives in building construction has been emphasized in a wide range of research. In general, the trade-off relationship between time and cost is well understood and there is ample research on the subject. However, despite sustainable building designs, relationships between time and environmental impact, as well as cost and environmental impact, have not been fully investigated. The objectives of this research were mainly to analyze and identify relationships of time, cost, and environmental impact, in terms of CO2 emissions, at different levels of a building: material level, component level, and building level, at the pre-use phase, including manufacturing and construction, and the relationships of life cycle cost and life cycle CO2 emissions at the usage phase. Additionally, this research aimed to develop a robust simulation-based multi-objective decision-support tool, called SimulEICon, which took construction data uncertainty into account, and was capable of incorporating life cycle assessment information to the decision-making process. The findings of this research supported the trade-off relationship between time and cost at different building levels. Moreover, the time and CO2 emissions relationship presented trade-off behavior at the pre-use phase. The results of the relationship between cost and CO2 emissions were interestingly proportional at the pre-use phase. The same pattern continually presented after the construction to the usage phase. Understanding the relationships between those objectives is a key in successfully planning and designing environmentally sustainable construction projects.

In this thesis, we present and analyze three algorithms that are designed to make computer simulation more efficient, valid, and/or applicable. The first algorithm uses simulation cloning to enhance efficiency in transient simulation. Traditional simulation cloning is a technique that shares some parts of the simulation results when simulating different scenarios. We apply this idea to transient simulation, where multiple replications are required to achieve statistical validity. Computational savings are achieved by sharing some parts of the simulation results among several replications. We improve the algorithm by inducing negative correlation to compensate for the (undesirable) positive correlation introduced by sharing some parts of the simulation. Then we identify how many replications should share the same data, and provide numerical results to analyze the performance of our approach. The second algorithm chooses a set of best systems when there are multiple candidate systems and multiple objectives. We provide three different formulations of correct selection of the Pareto optimal set, where a system is Pareto optimal if it is not inferior in all objectives compared to other competing systems. Then we present our Pareto selection algorithm and prove its validity for all three formulations. Finally, we provide numerical results aimed at understanding how well our algorithm performs in various settings. Finally, we discuss the estimation of input distributions when theoretical distributions do not provide a good fit to existing data. Our approach is to use a quasi-empirical distribution, which is a mixture of an empirical distribution and a distribution for the right tail. We describe an existing approach that involves an exponential tail distribution, and adapt the approach to incorporate a Pareto tail distribution and to use a different cutoff point between the empirical and tail distributions. Then, to measure the impact, we simulate a stable M/G/1 queue with a known inter-arrival and unknown service time distributions, and estimate the mean and tail probabilities of the waiting time in queue using the different approaches. The results suggest that if we know that the system is stable, and suspect that the tail of the service time distribution is not exponential, then a quasi-empirical distribution with a Pareto tail works well, but with a lower bound imposed on the tail index.

Many complex and interdependent systems engineering challenges involve more than one stakeholder or decision maker. These challenges, such as the definition and acquisition of future air mobility systems, are often found in situations where resources are finite, objectives are conflicting, constraints are restricting, and uncertainty in future outcomes prevail. Air mobility operational models which simulate fleet wide behavior effects over time, in various mission scenarios, and potentially over the entire design life-cycle, are always multi-dimensional, cover a large decision space, and require significant time to generate sufficient solutions to adequately describe the design space. This challenge is coupled with the fact that, in these highly integrated solutions or acquisitions, multiple stakeholders or decision makers are required to cooperate and reach agreement in selecting or defining the requirements for the design or solution and in its costly and lengthy implementation. However, since values, attitudes, and experiences are different for each decision maker, reaching consensus across the multiple criteria with different preferences and objectives is often a slow and highly convoluted process. In response to these common deficiencies and to provide quantitative analyses, this research investigates and proposes solutions to two challenges: 1) increase the speed at which operational solutions and associated requirements are generated and explored, and 2) systematize the group decision-making process, to both accelerate and improve decision making in these large operational problems requiring cooperation. The development of the Air Mobility Operations Design (AirMOD) model is proposed to address the first challenge by implementing and leveraging surrogate models of airlift capability across a wide scenario space. In addressing the second major challenge, the proposed Multi-Agent Consensus Reaching on the Objective Space (MACRO) methodology introduces a process to reduce the feasible decision space, by identifying regions of high probability of consensus reaching, using preference distributions, power relationships, and game-theoretic techniques. In a case study, the MACRO methodology is demonstrated on a large air mobility solution space generated by AirMOD to illustrate plausibility of the overall approach. AirMOD and MACRO offer considerable advantages over current methods to better define the operational design space and improve group decision-making processes requiring cooperation, respectively.

Un agent est un élément qui décide une action. Par ce formalisme très général on peut aussi bien désigner deux enfants jouant à pierre-papier-ciseaux, des êtres humains choisissant des produits sur un marché, un logiciel de routage calculant un plus court chemin sur Internet pour transporter des informations sur des routes numériques encombrées, qu’une enchère combinatoire automatique pour vendre des liens commerciaux et rapportant des milliards à google. Les chercheurs en théorie de la décision algorithmique et en théorie des jeux algorithmique – des mathématiciens et informaticiens – aiment à penser que ces exemples concrets peuvent être modélisés au moyen de systèmes décisionnels rationnels, aussi complexe la réalité soit-elle. Les systèmes décisionnels modernes trouvent leur complexité dans plusieurs dimensions. D’une part, les préférences d’un agent peuvent être complexes à représenter avec de simples nombres réels, alors que de multiples objectifs conflictuels interviennent dans chaque décision. D’une autre part, les interactions entre agents font que les récompenses de chacun dépendent des actions de tous, rendant difficile la prédiction des actions individualistes résultantes. L’objet de cette thèse en théorie algorithmique des systèmes décisionnels interactifs (jeux) est de poursuivre des efforts de recherche menés sur ces deux sources de complexité, et in fine, de considérer les deux complexités dans un même modèle. / An agent is an entity that decides an action. By using this abstraction, it is possible to model two children playing rock-paper-scissors, a software computing a shortest path on the internet for packet-routing on congest numerical networks, as well as an automatic combinatorial auction that sells commercial links in order to make google earn billions. The researchers in algorithmic decision theory and algorithmic game theory (mathematicians and computer scientists) like to think that these real-life examples can be modelled by mean of agents in an interaction decision system, no matter how complex is reality. The modern interactive decision systems find their complexity in multiple aspects. Firstly, the preferences of an agent can be complex to model with real numbers when there are multiple conflicting objectives resulting from every decision. Secondly, the interactions between agents are such that the payoff of every individual depends of the actions of all, making difficult the prediction of the resulting action-profile. This thesis aims at pursuing research efforts lead on these two sources of complexity, in order to consider ultimately both aspects in the same model.

The proposition of this dissertation is that superior capital budgeting solutions can be attained by not only analyzing projects individually but rather as part of a portfolio of projects that has the objective of maximizing the company’s range of multiple objectives, not only the economic benefit. The dissertation starts with a detailed study of current techniques and an assessment of flaws and shortcomings. This study concludes with the requirements that any new approach or model must address in order to improve on the current practices. Based on these requirements, a new model is developed based on the portfolio approach that integrates all the assumptions, constraints, project and variable interrelationships. An important feature of the model is that it selects its portfolio of capital projects in such a way that it optimizes support for the company’s multiple objectives, not only the economic objective. The dissertation concludes with the application of this model to a hypothetical case. It is concluded that, by developing and using this model, a company can improve the analysis required before capital budgets are finalized. / Dissertation (MEng (Industrial Engineering))--University of Pretoria, 2006. / Industrial and Systems Engineering / unrestricted

在電子商務的環境中，透過資訊科技的使用與通訊網路的連結，將會有愈來愈多的產品或服務需求者透過新興的電子化媒體 (如網際網路) 來尋找可行的交易互動夥伴，進而完成交易。因此，交易結構□每一份子間的互動關係，將面臨新的衝擊與挑戰。而納入電子商務觀念的供應鏈管理，將是以資訊科技與通訊技術為基礎的新領域，在此一領域中，供應鏈可以簡單地概念化成三部份：即產品/服務的供給者 (賣方)、產品/服務的需求者或是消費者 (買方)、及提供兩者溝通服務的資訊服務提供者。而在三者間，除了存在著生產與配送過程中既有的物料流/產品流、服務流及完成交易所必須的金流之外，更重要的是還有提供控制機能的資訊流。如何有效地管理與利用資訊流便成為供應鏈管理成功與否的關鍵性因素，而企業也因此產生了對於資訊服務的需求。 因此，本論文的目的在於發展出一個完整的研究體系，以針對傳統供應鏈中之供需雙方與資訊服務業之間的關係，發展出一個供需互動模式，以使電子商務環境中資訊服務的供需雙方能夠據此制定重要的決策與策略。此一體系包含了以下三個子體系：概念體系、評估體系、以及規劃體系。在概念體系的部份，本研究將透過文獻探討，針對供應鏈中的供需者 (可被視為資訊服務的需求者) 與資訊服務提供者，發展出一個整合的概念性互動模式，此一模式將解釋各個體之目標與其行為屬性，而這些目標與屬性也將成為後續評估及規劃體系發展的基礎。而後續兩個體系的發展，將以資訊服務的供需互動為研究主體。就評估體系而言，本研究將分別使用加法型 (層級分析法) 與非加法型 (模糊積分法) 方法來發展評選資訊服務提供者的多準則決策模式。而根據上述的結果，決策者便可針對其手邊現有的可選擇方案，來進行評選。一旦評選結果確定之後，決策者便可與其進行後續的供需互動。至於規劃體系的部份，則是要分析供需雙方如何根據自身的目標與資源限制，經由資訊的分享與交換，與所選取的夥伴進行互動。根據供需關係的型態及供需互動的主導者這兩個分類的標準，本研究將供需互動分成四種不同的狀況來探討。而透過模糊二階多目標規劃模式與多階段解題流程圖的應用，我們可以分析供需單位間如何透過資訊的交換以進行互動，並解釋互動所可能出現的結果，亦即失敗或成功。最後，本研究也將使用一個簡例來說明模式的可用性。 第一章 緒論…………………………………………… 1 第一節 研究動機與背景………………………… 1 第二節 研究目的………………………………… 3 第三節 研究方法與發展流程…………………… 5 第四節 論文結構與內容………………………… 6 第二章 文獻探討……………………………………… 7 第一節 電子商務………………………………… 7 壹、電子商務之定義……………………………. 7 貳、電子市場……………………………………. 12 第二節 供應鏈管理……………………………… 15 壹、供應鏈管理之定義………………………… 15 貳、關係的管理與分析………………………...… 17 參、買賣雙方之供需關係………………………... 21 肆、資訊服務提供者之中介……………………... 24 第三節 個體之目標與行為……………………… 29 壹、供應鏈管理之整體目標……………………... 29 貳、供給者 (賣方) 之立場……………………… 32 參、需求者 (買方) 之立場……………………… 35 肆、資訊服務提供者之立場…………………… 39 第三章 研究模式與方法……………………………… 49 第一節 研究模式………………………………. 49 壹、研究定位與個體定義………………………. 49 貳、供需互動模式………………………………. 51 參、研究範圍與分類架構………………………. 52 第二節 研究類型與步驟………………………… 54 第三節 評估方法論……………………………… 58 壹、因子分析……………………………………... 58 貳、加法型多準則評估…………………………. 59 參、非加法型多準則評估………………………. 61 肆、方案績效值的取得………………………… 63 第四節 規劃方法論……………………………… 70 壹、多目標規劃法…………………………… 70 貳、二階規劃法…………………………………. 73 第四章 評估面之研究 – 資訊服務提供者之評選…… 78 第一節 樣本特徵與資訊服務之使用現況……… 78 第二節 評選資訊服務提供者之準則分析……… 81 壹、評選準則之敘述統計分析………………… 81 貳、評選準則之因子分析……………………… 82 參、後續之效度驗證程序……………………… 90 第三節 多準則評估與決策體系之建立………… 93 壹、加法型多準則評估 – 層級分析法………… 93 貳、非加法型多準則評估 – 模糊積分法……… 97 參、實例說明與比較……………………………. 99 第五章 規劃面之研究 – 供需互動模式之發展…...….. 103 第一節 各種供需互動之說明…………………. 103 第二節 供需互動模式之發展………………… 106 壹、問題特性與解題流程……………………… 106 貳、互動規劃模式之建立……………………… 107 參、不同關係型態對互動過程的影響………… 113 第三節 簡例說明……………………………… 117 壹、背景說明…………………………………… 117 貳、問題求解過程說明………………………… 118 參、討論………………………………………… 125 第六章 結論與建議…………………………………… 127 第一節 結論……………………………………… 127 第二節 研究限制與困難………………………… 129 第三節 未來發展方向…………………………… 130 參考文獻………………………………………………… 131 附錄一………………………………………………………… 141 附錄二………………………………………………………… 145 附錄三………………………………………………………… 150 附錄四………………………………………………………… 153 附錄五………………………………………………………… 155 附錄六………………………………………………………… 163 附錄七………………………………………………………… 165 博士候選人簡歷……………………………………………… 172 / In the environment of Electronic Commerce (EC), there are more and more demanders of products or services looking for available interactive partners of transaction through the burgeoned electronic media (such as the Internet), who then complete transactions with the use of information technology and the connection of communication networks. Therefore, the interactive relationship between each member in the transaction structure will face new poundings and challenges. And the supply chain (SC) management, which fits into the notion of EC, will be a new field based on information technology and communication infrastructure. Within this field, the SC can be simply conceptualized into three parts: (1) Those act as the suppliers of products and services (the sellers), (2) The demanders or consumers of products and services (the buyers) and (3) the information service provider (ISP) which provides the information service for both parties. Among these three parties, in addition to the material/product flow and service flow existed in the production and distribution processes together with the financial flow required of accomplishing transactions, what is more important is the information flow that provides control function. Thus, how to effectively manage and use information flow becomes a key factor for successful SC management. As a result, the needs from enterprises for information service arise. This dissertation aims to establish a complete research system which helps develop an interactive supply-demand model for SC in EC, especially focusing on the relationship between the demanders and suppliers of information service. The research system includes three sub-systems: system of conceptualization, system of evaluation and system of planning. The system of conceptualization develops an integrated conceptual model to depict the interactive supply-demand relationship within SC. This model explains the objectives and the behavioral attributes of every individual, which then become the foundation of follow-up development of the systems of evaluation and planning. As for system of evaluation, this paper uses both additive (Analytic Hierarchy Process) and non-additive methods (Fuzzy Integral) to develop the multiple criteria decision making model for evaluating and selecting ISPs. In accordance with the results above, decision-makers are able to evaluate and select from alternatives on hand. Once the evaluation result is confirmed, decision-makers can proceed with the follow-up supply-demand interaction. As for the planning system, analysis of how supplier and demander of information service interact with each other according to their objectives and resource constraints is carried out. This dissertation also divides the supply-demand interaction into four different situations according to the type of relationship and the dominance. Through the application of fuzzy bi-level multiple objective programming (fuzzy BLMOP) technique and the multi-stage problem solving flow chart, we can analyze how the supply and demand units interact with each other by exchanging information and the possible outcomes of interactions can be explained. Finally, this dissertation illustrates the applicability of the fuzzy BLMOP model with a simple example.