Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Topics - Bertrand Mareschal

Pages: 1 ... 3 4 [5]
61
Jean Baptiste Rakotoarivelo / Abstract
« on: April 08, 2020, 03:28:45 PM »
La méthode PROMETHEE pour l'aide à la décision multicritère :
l'informatisation du principe des réalismes aux résultats des consultations par le
SIKIDY KARATA typiquement Malgache avec prescription des médicaments y
afférents. Cas de la région Boeny Mahajanga Madagascar

Résumé
Le sikidy1 est une pratique traditionnelle et coutumière. Jusqu'à nos jours, cette coutume est
encore héritée par des générations à Madagascar. L’utilisation de Sikidy est florissante dans
toutes les régions à Madagascar. Cependant, notre recherche se focalise dans la régions Boeny
sis au Nord-Ouest de grande île. Dans cette région la façon de consulter l'ombiasy2 est devenue
une croyance habituelle. Les résultats sont palpables et réellement prouvés, pour ne citer que le
traitement de l'angorongosy (jaunisse), qui s'avère les soins satisfaisant face à une fracture
osseuse suite à un accident ne nécessitant pas forcement une intervention médicale. Ainsi, la
particularité de sikidy karata3 typiquement malgache nous a incité à faire une nouvelle
invention afin de s’adapter au développement de la nouvelle technologie de l'information et de
la communication. Par ailleurs, nous utilisons la méthode d'aide multicritère à la décision
"PROMETHEE" afin de résoudre le problème exposé par le patient, et aussi être en mesure de
le rendre pratique pour les individus qui l'utilisent en fonction de leurs besoins concourant au
médicament associé. Les 32 cartes sont utilisées pour permettre à la connaissance des 28 cas
différents qui sont souvent associés à la vie quotidienne dans le faubourg Malgache. Dans ce
cas, Il existe plusieurs façons d'établir le rangement de ces 32 cartes. Mais d'après le code donné
pour chaque couleur, l’idée fondamentale restera toujours la même si les cartes changent de
disposition sur la table. Cette recherche s'inscrit dans l'idée de valoriser le monde de
l’anthropologie conjuguée à notre travail à Madagascar, voire à l'échelle mondiale, afin que les
usagers puissent savoir s'approprié du principe du réalisme par le biais de l'évènement et la
situation qui y sont rattachés.
Mots clés : le Sikidy, tradition nord-ouest de Madagascar, Anthropologie malgache, pratique
divination à Madagascar, technique divinatoire, coutumes et tradition malgache.
1 La divination malgache, appelée sikidy, est une pratique en usage sur tout le territoire de Madagascar, dont les
principes sont directement empruntés à la géomancie arabe. Auparavant la divination sikidy consiste à disposer
sur le sol des graines de fano (une sorte d'acacia), sous la forme d'un tableau, dans le but de lire la destinée à travers
certaines configurations de graine. C'est pourquoi qu'il est très difficile de coder en informatique à cause des
graines purement identique
2 Les devins sont appelés mpisikidy (« celui qui pratique le sikidy »), ou ombiasy, c’est-à-dire guérisseurs, et ils
sont des spécialistes reconnus par la société, leurs compétences étant le fruit d’une assimilation de techniques
complexes. Marcia Ascher, chapitre VI "Divination" 2000, p66
3 Karata : appellation particulière de la "carte" dans la région Boeny au Nord-Ouest de Madagascar, jeux de 32
cartes

62
Anastasia Blouchoutzi / Abstract
« on: April 08, 2020, 03:27:50 PM »
A multi-criteria decision analysis on the distribution of migrants’ accommodation facilities in Greece.
Anastasia Blouchoutzi
Dimitra Manou
Jason Papathanasiou
This paper is prepared in the framework of the H2020 project MAGYC (MigrAtion Governance and AsYlum Crises).
Greece has been in the frontline of the recent migrant and refugee crisis which occurred in a period of deep recession for the country. Moreover, Greece didn’t use to be a traditional destination for immigrants. Thus, the national and local authorities were unprepared to respond quickly and effectively to the asymmetrical pressure of the large scale arrivals. However, towards achieving the ultimate target of a sustainable integration policy for immigrants and refugees in their host countries, decision makers need to respond effectively in several kinds of challenges.
One of the long-term challenges for the policy makers in the migrants’ host countries is the optimal geographical distribution of migrants’ accommodation facilities so as to strengthen their integration outcomes and serve the crucial goal of social inclusion. The housing shortage for example usually prevents from the optimal allocation of the immigrants and the refugees based on the proximity of available job vacancies that could enhance the prospects for their successful labor market integration. Still, the government’s economically rational response to the immigrant inflows is their effective labor market integration. Almost five years after the outburst of the migration crisis, there is still an ongoing political debate in Greece on the appropriate placement policy of the immigrant and refugee population arriving in the country.
This paper focuses on the evaluation of the internal redistribution policy of immigrants and refugees in Greece as described by the current reception and accommodation scheme implemented in the country. Our main concern is to check whether broader socio-economic phenomena ranging from the access to labor market and education to the concentration of social networks have been taken into consideration by the national authorities while planning and implementing the dispersion policies. Furthermore, we intend to deepen into the outcomes of such policies with regard to the concerns arisen by the local authorities and communities.
Combining indicative indicators for the aforementioned areas of interest, including some of the EU Zaragoza Indicators for Immigrants Integration, data will be processed with the well-known multi-criteria decision analysis method PROMETHEE. In particular, the PROMETHEE method is going to provide us with a ranking of the Greek regions (NUTS 2 classification) following which we will evaluate the current allocation of migrants’ accommodation facilities. PROMETHEE belongs to the outranking family of multi-criteria methods and was chosen due to its relative simplicity and the robustness of the results. Using the method terminology, its steps include the pairwise comparison of actions for each of the chosen criteria, then the computation of unicriterion flows, and finally the aggregation of the latter into global flows.
This paper extends the literature on the immigrants’ and refugees’ integration in their host countries as well as the governance of migration. Specifically, the outcomes of our research will be used to unveil the effectiveness of the implemented immigrants’ and refugees’ allocation policies in Greece, compare it with other European countries’ immigrants’ and refugees’ placement practices, illuminate the reasoning for the current settlement and provide the decision makers with policy suggestions on alternative allocation policies.

63
Marco A. Faveri / Abstract
« on: April 08, 2020, 03:23:25 PM »
MCDA to improve resources distribution on criminal investigations at Brazilian Federal Police: sorting crime reports in importance and urgency categories
This study presents an MCDA model to deal with crime reports that arrives at the many units of Brazilian Federal Police, using FlowSort methods (crispy, interval and fuzzy). The model aggregates several criteria in two dimensions, one of importance and another of urgency. The importance dimension takes into account seven criteria, like minimum and maximum Amount of Imprisonment penalty, Complexity of Crime, Value of Economic Injury, Hierarchy of Legal Goods and Days After the Facts. In the other side, to characterize the urgency dimension were utilized five criteria, like the Moment of Offense Commitment, Situation of Crime Product, Situation of Crime Evidences, Suspect Behavior and Punitive Prescription. The FlowSort method is then applied to assist in the distribution of crime reports into ordered categories of importance and in another ordered categories of urgency. Both dimensions are considered simultaneously to analyze criminal reports. As a case study, fifteen real criminal reports were chosen and submitted to the model, approved by most consulted experts (89.1% for importance and 83.2% for urgency approval rates). Extending the research and in order to improve the analysis of each Criminal Report when there is a higher level of uncertainty, two other models for importance sorting were created using Interval FlowSort and Fuzzy FlowSort methods. Applying these models, it was possible to observe more differences between the actions, exploring the imprecision and the uncertainty, preserving the decision maker’s subjective information, that was not possible using the traditional FlowSort method. Wherefore, the model supports appropriate use of public resources, aligned with the objective of the Brazilian Federal Police and required from National Constitutional Law, by highlighting more important and urgent investigations. The introduction and adoption of these techniques on the organization are gradually occurring. Moreover, because the number of investigations, policemen and police units, the wide dissemination of MCDA methods also depends on computational resources availability. Hence, the developing of specifics applications are required to aggregate the necessary information and to report the decision maker´s recommendations.

64
Giselle Tavora / Abstract
« on: April 08, 2020, 03:05:29 PM »
Elicitation of Criteria Importance Weights through the Simos Methods – A Case Study on Decommissioning Subsea Structures
Giselle da Silva Távora, Alberto Luiz Coimbra Institute – Graduate School and Research in Engineering, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil, giselletavora@oceanica.ufrj.br
Carlos Eduardo Infante, Federal University of São João del Rei, prof.eduinfante@gmail.com
Laurelena Crescencio Palhano, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil, laure@sage.coppe.ufrj.br
Marcelo Igor Lourenco de Souza, Alberto Luiz Coimbra Institute – Graduate School and Research in Engineering, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil, migor@lts.coppe.ufrj.br
Jean-David Caprace, Alberto Luiz Coimbra Institute – Graduate School and Research in Engineering, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil, jdcaprace@oceanica.ufrj.br
With the end of life of several Brazilian oil fields, the decommissioning of submarine structures is a challenge to be faced, since there are still no consolidated methodologies to assess the general view of the problem to be found. Currently, pipelines represent a significant amount of all structures to be removed and the final disposition is one of the most important aspects to be analyzed. Decommissioning is expected to occur, once the depletion of resources or the useful life of the production units are reached.
Total removal by cutting and elevating sections of pipe, removal of the entire length of the pipe by reverse s-lay, abandonment in situ with dumping rocks at the ends of the pipe and total abandonment are several alternatives to be considered in the subsea decomissinging problem. Meawhile, the creiteria to be considered for the assessment are: Environment, Safety, Economy, Tehcnical and Social.
A significant factor pertaining to the non-compensatory multicriteria decision aiding models (MCDA), such as the outranking methods, i.e. PROMETHEE, is the criteria weighting, or the importance of the criteria. Generally, these parameters imprint the preferences of a single decision maker to the model. The existing methods, which are widely used to assess the criteria importance weights, could be classified into two categories: direct assessment procedures, where the decision maker is asked to explicitly express the criteria weights in terms of percentages, and indirect methods, inferring the weights from pairwise comparisons of the criteria.
The method proposed by Jean Simos in 1990 has gained popularity and has been applied to different types of problems to solve the weaknesses of the previous methods, due to its simplicity, and the convenience it provides to the decision makers to express his preferences.
Specifically, it requires the construction of a hierarchy on the evaluation criteria, by involving the decision maker to a “playing cards” procedure, in order to attribute numerical values to them. Nevertheless, the process recommended by Simos and its revised version proposed by Figueira and Roy (2002) have some robustness issues. In particular, they arbitrarily calculate a unique weighting vector, even though there exist infinitely more weight vectors, also satisfying the preferential statements, which have been defined by the DM during the initial arrangement of the cards.
The purpose of this paper is to expose the robustness problem of the improved Simos method observed during a large scale case study applied to decommissioning of subsea structures.
The results has been collected during a workshop held with the presence of several Oil and Gas companies, educational institutions and regulatory bodies. The participants were divided into five people per table, each participant specialized in one criterion. The people around the table and the scenarios to be analysed has been systematically varied in order to analise the influence of each decision maker teams.
In each scenario, characteristics related to distance from the oil field, water depth, presence or absence of diving, location of the country, presence or absence of Naturally-Occurring Radioactive Materials, crossings of pipes, type of seabed, type of fishing and marine life were proposed.
The main outcome of the study shows the variance between the different weights proposed by the decision making teams depending of the people around the table and the scenarios considered. The robustness problem of the Simos method has therefore been exposed and quantified for this specific situation.

65
Dimitra Zangana / Abstract
« on: April 08, 2020, 03:04:13 PM »
Energy Design and Multicriteria Decision Analysis in a New Residential Building - Study on the Application of Thermal Insulation and Systems
Abstract
In this study, an attempt was made to design and study a new two-storey house with energy planning criteria. At the theoretical level, the principles and systems used for building energy planning and the importance of multicriteria analysis in decision-making on this issue were analyzed.
The main criterion for the energy efficiency of the construction outside the bioclimatic design was the thermal insulation of the building where multi-criteria analysis was applied with the Visual Promethee application to select the most suitable thermal insulation material. Equal weights of criteria were analyzed, but also different weights calculated through personal interviews with civil engineers were applied. The results of the two analyses were compared. In the case of analysis with different criteria weights, the expanded polystyrene is followed by stone wool, while for the analysis with similar criteria weights expanded cork ICB is the best alternative.
Also, based on the case with similar weights, the two predominant thermal insulation materials (expanded polystyrene and rock wool) were introduced into the building's components and analyses were carried out showing the needs for cooling and heating of the new home in the four climate environments using Revit. Comparison and annotation of the analyses for the heating and cooling needs according to the peak loads for the two thermal insulation materials in the climate zones also took place.
In terms of peak load for cooling, rockwool in climatic zones B and D exhibits little difference in efficacy in comparison to expanded polystyrene. In the other two climate zones expanded polystyrene predominates. Polystyrene is particularly effective when applied to climatic zone A. On the other hand, for peak loads, rockwool in climatic zone B and D. performs better in climatic zone B than in climatic zone B compared to expanded polystyrene. while in zone D the largest difference is found in the two materials. In the other two zones, expanded polystyrene is superior to rock wool, with a significant difference being present in climate zone A. On the ground floor, climate zones A and C are more efficient, while for the other two, rock wool is the most efficient. On floor for climatic zones C and D the most effective is cotton wool, for climatic zone A expanded polystyrene while for B the insulating materials exhibit similar peak loads for cooling. Finally, the building is presented with proposed interventions such as the addition of passive and energy solar systems, addition of vegetation and devices that enhance natural cooling / lighting.
Keywords: Multi Criteria Analysis, Energy Design, Peak Loads, Visual PROMETHEE, Revit.

66
Bertrand Mareschal / Abstract
« on: April 08, 2020, 03:01:35 PM »
How to choose the correct preference function with PROMETHEE
The PROMETHEE methods are one of the most widely used families of outranking methods. They rely on preference modeling through the definition of preference functions and the assessment of different thresholds. This step is often overlooked or done in an inappropriate way. This can lead to wrong results. In this paper, we recall the definition of the PROMETHEE preference functions and we introduce general guidelines to properly assess preference functions to different types of criteria and to define the values of the associated thresholds

67
Jean-Philippe Waaub / Abstract
« on: April 08, 2020, 03:00:58 PM »
Participation of stakeholders in MCDA: processes and tools
This contribution emphasizes the participation of stakeholders in the field of environmental assessment and more specifically their participation supported by a multi-criteria decision aid approach (MCDA) which could also be presented as a multicriteria aiding approach for consultation, deliberation or negotiation in view of a decision. In this regard, the system of actors for environmental assessment is presented. The processes involved must be aligned: planning, environmental assessment, participation, multi-criteria support. Environmental assessment is a complex field, where interdisciplinarity is a necessity which can judiciously be handled by a multi-criteria approach, where the stakeholders are potentially very numerous, and where the search for solutions and the related issues are often conflicting. The necessary prerequisites to implement a participatory and contributory approach are therefore specified. Stakeholder participation can take different forms depending on the MCDA methods that support environmental assessment. It can be more or less contributory depending on the degree of stakeholder involvement and also depending on the stages in which stakeholders are invited to participate. It can also be supported in a flexible and non-directive manner or even be more or less strongly framed. Depending on the resources available, participation can be facilitated by an expert member of the environmental assessment team or benefit from a facilitator, or even a mediator or negotiator. The advantages and disadvantages of insisting on the settlement of certain disagreements or conflicts upstream of the process or more downstream are presented. The emphasis is on sharing upstream the performance table of the scenarios analyzed or rather to postpone these questions at the end of the process, when providing the recommendations to the authority in charge of making the final decision to accept or not and under what conditions the proposal submitted for environmental assessment. Each step of the environmental assessment is discussed with regard to stakeholder participation, supported by a MCDA approach and the tools that can support that participation. From here, all the elements are treated within the framework of the PROMETHEE and GAIA multi-criteria methods implemented by the Visual PROMETHEE software. Of course, many elements are also applicable following other multi-criteria methods. The very first thus addresses the question of who participates and at what stages depending on the problem posed either at the project or strategic level (policy, plan, program). The concept of restricted societal participation is discussed. With regard to the stages of problem definition, namely: the identification of scenarios, issues, criteria and measurement indicators, collective working tools are proposed. With regard to the scenario performance table according to each of the impact criteria, the question of scientific knowledge, and local, contextualized, endogenous or traditional knowledge is addressed. Sometimes you also have to deal with data gaps or missing data. Proposals are made to take advantage of the available experts as well as the stakeholders involved. Regarding stakeholder preferences and priorities, a few tools are available. It is indeed not easy for the participants to establish the bases for comparing the scenarios on the same criterion according to the measurement scale used. They do not all have the same way of setting their preferences. Also, how to help them establishing their priorities among the issues and the criteria that reflect them. Should the criteria be weighted or not? If yes, how to proceed? Finally, the tools offered by Visual PROMETHEE help to facilitate discussions between participants. The contributions of each of these tools are presented, including the sensitivity and robustness analyzes essential to deal with indecisions, uncertainties or specific conflicts. The questions related to the sharing of results specific to each participant and to the group are discussed, as well as those related to the degree of interactivity of the participants with the Visual PROMETHEE software is also presented. All of these points are illustrated from case studies conducted by the author and the GEIGER team over the past 20 years.

68
He Huang / Abstract
« on: April 08, 2020, 03:00:02 PM »
PROMETHEE V in Multi-Stakeholders – Reaching the Consensus on portfolios under Constraints
Decisions including public stakeholders are often complex and need to consider various conflicting criteria. PROMETHEE is an outranking method to treat multi-criteria problems with a discrete set of available alternatives that has been widely applied in public decision making. In order to find a preferred alternative, PROMETHEE I and II can be used to determine a partial or complete ranking based on the decision-makers' (DMs) preference. Some real-world decisions, however, require determining a set or portfolio of alternatives regarding goals and constraints given by multiple DM. Multi-Actor Multi-Criteria Analysis (MAMCA) is applied to cluster the points of view from DMs with different interests and preferences, so-called stakeholders. The goal is to find a subset of good alternatives maximizing the PROMETHEE net flow while reaching consensus among the involved DMs. We present a linear optimization model based on PROMETHEE V to reach consensus between the DMs while considering various constraints. We illustrate this approach with a case study of a school street implementation case from Learning Loops in the Public Realm (LOOPER) project which is aimed to develop new ways of decision-making that bring together citizens, stakeholders, and policy-makers by building a participatory co-creation methodology and platform that demonstrate ‘learning loops’. Different alternatives are proposed to improve safety around a school located in Schaerbeek. More than one alternative can be chosen to achieve the objective. However, there are some constraints for these alternatives, e.g., budget for the infrastructures and limited labor powers. By applying the model, sets of alternatives are chosen that satisfy the stakeholders and the given constraints.

69
Zhor Chergui / Abstract
« on: April 08, 2020, 02:58:56 PM »
Promethee methods: sensibility study & remarks
In the case of a decision rule based on an aggregation function that places us in the Arrow’s theorem
context, it is not possible to check some mathematical properties simultaneously. Indeed, to
construct a method using the same concepts, we can not abandon the principles of unanimity,
universality and non-dictatorship. Thus, the only two principles between which we have to choose
are the transitivity and the independence. In other words, it is impossible to build an ordinal method
verifying the transitivity and the independence at the same time. A thorough analysis has shown the
impossibility to find and\or define MCDM methods satisfying some derived mathematics properties
simultaneously.
In this paper, we study the sensibility of Promethee family methods to the use of different versions
of independence and transitivity. On this basis, we construct rules and mathematical conditions
upon which Promethee family keep their original results. In case of change of the original
outranking, we propose post-optimality studies and enquiries allowing to expect the new results and
their values.
MCDM methods; performance evaluation; preference modelling; multicriteria analysis; Promethee;
independence property; transitivity property.
References
[1] Brans, J. P.& Mareschal, B. (1997) How to decide with PROMETHEE, ULB and VUB Brussels Free
Universities, Research repport.
[2] Brans, J. P.& Mareschal, B. (2001) PROMETHEE-Gaia : une méthodologie d’aide à la décision en présence
de critères multiples, éditions de l’université de Bruxelles, Ellipses.
[3] Doignon, J.-P., Monjardet, B., Roubens, M. & Ph. Vincke, (1986) Biorder families,valued relations, and
preference modelling, Elsevier Science : Journal of Mathematical Psychology, vol.30(4), p.435–480.
[4] Pirlot, M. and Vincke, P. (1997) Semiorders : Properties, Representations, Applications, Springer Netherlands
: Theory and Decision Library B.
[5] Vincke, Ph., (1982) Arrow’s theorem is not a surprising result, Elsevier Science : Europeanjournal of
operational research, vol.10(1), p.22–25.
[6] Vincke, Ph. (1992) Exploitation of a crisp relation in a ranking problem, Springer US : Theory and Decision,
vol.32(3), p.221–240.

70
General Discussion / Let us stick together, virtually
« on: March 26, 2020, 01:17:10 PM »
Dear forum members, Dear friends,

We are all facing difficult times that have a strong impact on our personal and professional lives. And we don't know how the situation will evolve during the next weeks/months.

I sincerely hope that all of you, your families and your friends, are well. It would be nice for all the members of our community to know that you are safe.

Please don't hesitate to share information and news through the PROMETHEE Methods LinkedIn group:

https://www.linkedin.com/groups/3956394/

I will be very happy and reassured to receive news from you.

With respect to the organization of PD2020 (which is much less important than our health), the Scientific and Organizing Committees are waiting until the end of the month to reassess the situation.

Please take care.

Best regards,
Bertrand

Pages: 1 ... 3 4 [5]