In several types of embedded systems the applications are deployed both as software and as hardware components. For such systems, the partitioning decision is highly important since the implementation in software or hardware heavily influences the system properties. In the industry, it is rather common practice to take deployment decisions in an early stage of the design phase and based on a limited number of aspects. Often such decisions are taken based on hardware and software designers‟ expertise and do not account the requirements of the entire system and the project and business development constraints. This approach leads to several disadvantages such as redesign, interruption, etc. In this scenario, we see the need of approaching the partitioning process from a multiple decision perspective. As a consequence, we start by presenting an analysis of the most important and popular Multiple Criteria Decision Analysis (MCDA) methods and tools. We also identify the key requirements on the partitioning process. Subsequently, we evaluate all of the MCDA methods and tools with respect to the key partitioning requirements. By using the key partitioning requirements the methods and tools that the best suits the partitioning are selected. Finally, we propose two MCDA-based partitioning processes and validate their feasibility thorough an industrial case study.

Discrete choice experiments (DCEs) are widely used across economic disciplines to value multi-attribute commodities. DCEs ask survey-respondents to choose between mutually exclusive hypothetical alternatives that are described by a set of common attributes. The analysis of DCE data assumes that respondents consider and trade all attributes before making these choices. However, several studies show that many respondents ignore attributes. Respondents might choose not to consider all attributes to simplify choices or as a preference, because some attributes are not important to them. However, empirical approaches that account for attribute non-consideration only assume simplifying choice behaviour. This thesis shows that this assumption may lead to misleading welfare conclusions and therefore suboptimal policy advice. The analysis explores 'why' attribute are ignored using statistical analysis or by asking respondents. Both approaches are commonly used to identify attribute non-consideration in DCEs. However, the results of this thesis suggest that respondents struggle to recall ignored attributes and their reasons for non-consideration unless attributes are ignored due to non-valuation. This questions the validity of approaches in the literature that rely on respondents' ability to reflect on their decision rule. Further analysis explores how the complexity of choices affects the probability that respondents do not consider all attributes. The results show that attribute consideration first increases and then decreases with complexity. This raises questions about the optimal design complexity of DCEs. The overall findings of the thesis challenge the applicability of current approaches that account for attribute non-consideration in DCEs to policy analysis and emphasis the need for further research in this area.

Best practice in gifted and talented identification procedures involves making decisions on the basis of multiple measures. However, very little research has investigated the impact of different methods of combining multiple measures. This article examines the consequences of the conjunctive ("and"), disjunctive/complementary ("or"), and compensatory ("mean") models for combining scores from multiple assessments. It considers the impact of rule choice on the size of the student population, the ability heterogeneity of the identified students, and the psychometric performance of such systems. It also uses statistical simulation to examine the performance of the state of Georgia's mandated and complex multiple-criteria assessment system.

This article presents a conceptual benchmarking framework which applies a multiple criteria approach to assess performance. In the process, a multiple criteria procedure is used to assess the performance of three hundred and ninety two (392) Portuguese firms. Based on the results of this procedure, a conceptual framework is devised to facilitate addressing relevant benchmarking implications. The framework is designed to provide a conceptual linkage between the performance measurement methodology and the organizational benchmarking system.

L’aide multicritère à la décision (AMCD) vise à faciliter et améliorer la qualité du processus de prise de décision. Les méthodes d’AMCD permettent de traiter les problèmes de choix, rangement et classification. Ces méthodes impliquent généralement la construction d’un modèle. Déterminer les valeurs des paramètres de ces modèles n’est pas aisé. Les méthodes d’apprentissage indirectes permettent de simplifier cette tâche en apprenant les paramètres du modèle de décision à partir de jugements émis par un décideur tels que “l’alternative a est préférée à l’alternative b” ou “l’alternative a doit être classifiée dans la meilleure catégorie”. Les informations données par le décideur sont généralement parcimonieuses. Le modèle d’AMCD est appris au cours d’un processus interactif entre le décideur et l’analyste. L’analyste aide le décideur à formuler et revoir ses jugements si nécessaire. Le processus s’arrête une fois qu’un modèle satisfaisant les préférences du décideur a été trouvé. Le “preference learning” (PL) est un sous domaine du “machine learning” qui s’intéresse à l’apprentissage des préférences. Les algorithmes de ce domaine sont capables de traiter de grands jeux de données et sont validés au moyen de jeux de données artificiels et réels. Les jeux de données traités en PL sont généralement collectés de différentes sources et sont entachés de bruit.Contrairement à l’AMCD, il existe peu ou pas d’interaction avec l’utilisateur en PL. Le jeu de données fourni en entrée à l’algorithme est considéré comme un échantillon éventuellement bruité d’une “réalité” ou “vérité de terrain”. Les algorithmes utilisés dans ce domaine ont des propriétés statistiques fortes leur permettant de s’affranchir du bruit dans ces jeux de données. Dans cette thèse, nous développons des algorithmes d’apprentissage permettant d’apprendre lesparamètres de modèles d’AMCD. Plus précisément, nous développons une métaheuristique afin d’apprendre les paramètres d’un modèle appelé MR-Sort (“majority rule sorting”). Cette métaheuristique est testée sur des jeux de donnéesartificiels et réels utilisés dans le domaine du PL. Nous utilisons cet algorithme afin de traiter un problème concret dans le domaine médical. Ensuite nous modifions la métaheuristique afin d’apprendre les paramètres d’un modèle plus expressif appelé NCS (“non-compensatory sorting”). Finalement, nous développons un nouveau type de règle de veto pour les modèles MR-Sort et NCS qui permet de prendre les coalitions de critères en compte. La dernière partie de la thèse introduit les méthodes d’optimisation semi-définie positive (SDP) dans le contexte de l’aide multicritère à la décision. Précisément, nous utilisons l’optimisation SDP afin d’apprendre les paramètres d’un modèle de fonction de valeur additive. / Multiple-criteria decision analysis (MCDA) aims at providing support in order to make a decision. MCDA methods allow to handle choice, ranking and sorting problems. These methods usually involve the elicitation of models. Eliciting the parameters of these models is not trivial. Indirect elicitation methods simplify this task by learning the parameters of the decision model from preference statements issued by the decision maker (DM) such as “alternative a is preferred to alternative b” or “alternative a should be classified in the best category”. The information provided by the decision maker are usually parsimonious. The MCDA model is learned through an interactive process between the DM and the decision analyst. The analyst helps the DM to modify and revise his/her statements if needed. The process ends once a model satisfying the preferences of the DM is found. Preference learning (PL) is a subfield of machine learning which focuses on the elicitation of preferences. Algorithms in this subfield are able to deal with large data sets and are validated withartificial and real data sets. Data sets used in PL are usually collected from different sources and aresubject to noise. Unlike in MCDA, there is little or no interaction with the user in PL. The input data set is considered as a noisy sample of a “ground truth”. Algorithms used in this field have strong statistical properties that allow them to filter noise in the data sets.In this thesis, we develop learning algorithms to infer the parameters of MCDA models. Precisely, we develop a metaheuristic designed for learning the parameters of a MCDA sorting model called majority rule sorting (MR-Sort) model. This metaheuristic is assessed with artificial and real data sets issued from the PL field. We use the algorithm to deal with a real application in the medical domain. Then we modify the metaheuristic to learn the parameters of a more expressive model called the non-compensatory sorting (NCS) model. After that, we develop a new type of veto rule for MR-Sort and NCS models which allows to take criteria coalitions into account. The last part of the thesis introduces semidefinite programming (SDP) in the context of multiple-criteria decision analysis. We use SDP to learn the parameters of an additive value function model.

No description available.

No description available.

Thesis (MBA)--Stellenbosch University, 2003. / ENGLISH ABSTRACT: One of the most difficult and more important decisions taken by a client is the selection of the most appropriate contractor. It requires the assessment of various factors, often conflicting, in order to determine the most appropriate contractor and are therefore classified as a problem that can be resolved by using multiple criteria decision making methods. The act of decision making is never an easy one and requires a sound understanding of the requirement, the alternatives and the model used to assess the alternatives in terms of the requirement in order to instil confidence that the most appropriate alternative is selected. The appropriateness of the methods used in contractor evaluation has a vital impact on the cost of the transaction. The three broad categories, or schools of thought, relating to multiple criteria decision making (MCDM) methods are assessed in terms of their applicability to the contractor selection problem within a quasi-government organisation, namely Armscor. Of the three categories, only the methods of the value measurement category were found to be appropriate within the current legislative framework of the Preferential Procurement Act, which seeks to express the performance of an offer as a unique numerical function. The old contractor selection model of direct point allocation on a qualitative scale is shown to be inappropriate, especially in terms of the additive utility assumption of single dimensional units. The proposed new model makes use of the weighted product model that is not restricted by the additive utility assumption as it results in dimensionless analysis of the criteria. The utility functions associated with the quantitative criteria uses curves which are raised to the power of the confidence variable. The arithmetic mean of these variables represents the group’s confidence level associated with each contractor’s offer in the correctness and/or its ability to maintain the stated level of performance. Furthermore, the analytic hierarchy process is used for the assessment of the qualitative criteria. The new model, although not perfect, is an improvement over the old model with regards to the understanding of the requirement as well as the assessment of contractors’ proposals. / AFRIKAANSE OPSOMMING: Die keuse van ‘n kontrakteur is een van die moeilikste besluite wat ‘n kliënt moet neem, dit is egter ook een van die belangrikste besluite wat geneem word. Ten einde die mees geskikte kontrakteur te kies, moet daar ‘n waarde geheg word aan verskeie faktore, menigmaal teenstrydig, wat kontrakteur seleksie klassifiseer as ‘n probleem wat deur middel van meervoudige-kriteria-besluitnemingsmetodes opgelos kan word. Die handeling van besluitneming is nooit ‘n maklike een en vereis deeglike kennis van die behoefte, die alternatiewe, asook die model wat gebruik word om die alternatiewe in terme van die behoefte te waardeer in orde om vertroue in die gekose alternatief te hê. Vir die doeleindes van hierdie studie is die drie kategorieë van meervoudige-kriteria-besluitnemingsmetodes vergelyk in terme van hul toepaslikheid op die voorafgenoemde probleem binne ‘n Semi-Staatsinstelling, naamlik Krygkor, met die oogmerk om die beste metode te identifiseer. Slegs die metodes vervat in die waarde-meting kategorie is geskik binne die Wet op die Raamwerk vir Voorkeurverkrygingsbeleid wat die evaluasie van ‘n aanbod uitdruk as ‘n unieke numeriese funksie. Uit die studie blyk dit dat die vorige kontrakteur seleksie model van direkte punt allokasie op ‘n kwalitatiewe skaal onvanpas is, veral in terme van die sommerings-nutfunksie aanname van enkel dimensionele eenhede. Die model wat eerder aanbeveel word, maak gebruik van die geweegde-produk-model wat nie beperk word deur die bogenoemde aanname nie, aangesien dit dimensielose analise tot gevolg het. Nutfunksies wat geassosieër word met kwantitatiewe kriteria, word voorgestel deur kurwes wat tot die mag van die vertrouensvlak-veranderlike gehef word. Die rekenkundige gemiddelde van hierdie veranderlike verteenwoordig die groep se vertrouensvlak met betrekking tot elke kontrakteur se akkuraatheid en vermoeë om die gespesifiseerde vlak van werkverrigting te handhaaf. Die kwalitatiewe kriteria word beoordeel deur gebruik te maak van die analitiese hiërargie proses. Die gevolgtrekking wat uiteindelik gemaak word is dat die nuwe model, alhoewel nie foutloos, tog ‘n verbetering is op die vorige model, veral met betrekking tot die insig wat verkry word deur die ontleding van die kontrakteurs se voorstelle in terme van die bepaalde behoefte wat bevredig moet word, ten einde die beste keuse uit te oefen.

分析層級程序法在產品刪除決策之應用 / The recent explosion of product management in consumer packaged goods has highlighted the importance of product assortment decisions. In particular, firms are increasingly faced with the decision of which products to delete from distribution. Upon reflection, there are both strategic and tactical dimensions to this decision. Strategic approaches focus on the development of optimal product assortments as the basis for deletion decisions. Tactical approaches address incremental (i.e., item-by- item) decisions whether to delete any product, and if so, which product. This thesis focuses on tactical approaches and proposes using Analytic Hierarchy Process (AHP) as a systematic and analytic tool that helps to quantify the managerial judgments in identifying the candidates for product deletion. Supported by a practical case study, which illustrates how AHP can be beneficial in quantifying both financial and non-financial product performance rankings for managers’ easier understanding and higher transparency of product deletion decision-making.

Conflict analysis and game theory models are applied to a case study in the upper Rio Grande river basin. The objective is to find which theory best describes past developments in the Rio Grande river basin and the status quo of water use strategies employed by the players (decision makers). By assuming that these past properties will propagate in the future, the preferable change in the equilibrium solution is derived under climate fluctuation, coupled with future population growth scenarios. Past and future Rio Grande resource allocation conflicts are analyzed using (1) multicriterion decision making (MCDM) techniques, such as distance based approach of compromise programming and outranking technique of the ELECTRE family and (2) voting scheme approach of game theory. MCDM and game theory model cases are classified according to the following categories: 1. If decision makers consider each other payoff or if an authority above forces them to consider each other's payoffs, then the conflict analysis problem is a multiactor/ multiobjective problem. 2. If decision makers only care about their own payoff and not what other players payoff are, then the conflict analysis problem is described and solved by game theoretic models. Fifteen decision makers from the Rio Grande water allocation and water management conflict are used as an example to present the different approaches to conflict modeling. From the MCDM techniques used, namely the compromise programming of distance-based approach and the ELECTRE family of outranking relation, the former method stands out as being the most flexible and comprehensive methodology. Though these two methods are conceptually different, for this case study, both methods give approximately the same results. For the game theory analysis, the special voting scheme stands out as being the preferred approach because it better reflects the decision maker's preference and it also is easy to implement and apply. Finally, the climate change scenarios are considered, the 1XCO₂ and the 2XCO₂. Results obtained from these two scenarios indicate the Rio Grande river will face extreme water shortages that will require the development of a different set of water release rules.