Plenary Session – From qualitative assessments by experts and stakeholders to policy priorities: a multicriteria sorting approach

Luis C. Dias, Universidade de Coimbra, Faculdade de Economia e CeBER

In this talk, I will share a recent experience in eliciting qualitative assessments for the importance of the objectives and the impacts of policies, and then performing a multicriteria analysis based on those inputs. These assessments have been performed in the context of a European project on dealing with plastic litter in seas, involving participants from academia, public administration, and industry. I will discuss the aspects of aggregating the perspectives of multiple participants to obtain constraints on the criteria weights respecting the qualitative nature of their assessments, and their use to obtain robust conclusions. In this case the ELECTRE TRI sorting method was used, but similar strategies can be applied for group decision-making based on other multicriteria methods.

Short bio:

Luis C. Dias received his Ph.D. in Management Science from the University of Coimbra, where he is currently a Professor in Management Science and Director of the Centre for Business and Economics Research at the Faculty of Economics (FEUC), a collaborator at INESC Coimbra institute, and a member of the coordination board of Univ. Coimbra’s Energy for Sustainability Initiative. He is also currently the Chair of the General Assembly of the Portuguese Operational Research Society (APDIO), an Area Editor for the Omega journal and he is part of the editorial board of the EURO J. on Decision Processes and the Group Decision and Negotiation journal. In the past he has served as FEUC’s Vice-dean for Research, as a Vice-President of APDIO, and as a Subdirector of INESC Coimbra. He has been for brief periods a visiting professor at the University of Paris-Dauphine and the University of Vienna. Luis has published over a hundred articles in books and journals, including some of the main journals in the management science and operations research area, as well as some of the main journals in the energy and environment area. His research interests include multicriteria decision analysis, performance assessment, group decision and negotiation, and sustainability.


Tutorial session – A web-based APP for supporting Multicriteria Decision Modelling of outranking choice Problems

Helder Gomes Costa, (UFF)

Abstract: Types of decision problems and situations. Outranking decision situations and methods. Using a web-based app for modelling outranking decision problems: setting the problem, importing datasets, getting the kernel and the dominated subset, sensibility analysis regarding cut-levels and criteria constant of scales (or ‘weights’).

Short bio:

Graduation in Mechanical Engineering from Universidade Federal Fluminense (1987), master’s at Mechanical Engineering from Pontifícia Universidade Católica do Rio de Janeiro (1991) and doctorate at Mechanical Engineering from Pontifícia Universidade Católica do Rio de Janeiro (1994). He is currently full professor at Universidade Federal Fluminense. acting on the following subjects: decision multicriteria decision making. clustering and performance evaluation


Tutorial session – Applications of data-driven methods in multiple criteria decision making

Speaker: Leonardo Tomazeli Duarte (FCA/UNICAMP)

Abstract: The usual approach in multiple criteria decision making (MCDM) is based on the elicitation of preferences from decision makers. There are some situations, however, in which preference information is not available. In other words, it is not possible to perform elicitation processes, which means that tasks such as the estimation of the parameters of aggregation functions must be carried out by only considering data such as the decision matrix.  This  situation bears strong resemblance with a paradigm known as unsupervised learning, one of the current hot topics in machine learning. In this tutorial, we shall exploit the application of these data-driven (or unsupervised) approaches in MCDM. We will consider problems related to the estimation of the parameters of aggregation functions. We also discuss how these data-driven solutions can be equipped with strategies that ensure fair rankings in situations in which the alternatives are composed of two (or more) different groups.

Short bio:

Leonardo Tomazeli Duarte received the B.S. and M.Sc. degrees in electrical engineering from the University of Campinas (UNICAMP), Brazil, in 2004 and 2006, respectively, and the Ph.D. degree from the Grenoble Institute of Technology (Grenoble INP, Université Grenoble Alpes), France, in 2009. Since 2011, he has been with the School of Applied Sciences (FCA) at UNICAMP, Limeira, Brazil, where he is currently an associate professor. He is the Scientific Coordinator of the Track “Method” within the Brazilian Institute of Data Science (BI0S), one of the Brazilian Applied Research Center on Artificial Intelligence. He is also a member with the Laboratory of Data Analysis and Decision Aiding (LAD2/CPO) and with the Laboratory of Signal Processing for Communications (DSPCom lab). He is a Senior Member of the IEEE. In 2016, he was a Visiting Professor at the École de Génie Industriel (GI-Grenoble INP, France). Since 2015, he has been recipient of the National Council for Scientific and Technological Development (CNPq, Brazil) productivity research grant. In 2017, he was the recipient of UNICAMP “Zeferino Vaz” Academic Recognition Award (for research and teaching performance at UNICAMP). Her research interests center around the broad area of data science and lie primarily in the fields of signal processing, decision aiding and machine learning, and also in the interplays between these fields.