Hybridizing Multi-Objective Interactive Techniques - Some Applications to the Electricity Generation Industry
Francisco Ruiz, University of Málaga, Spain
Online Strategies for Hard Optimization Problems in Graphs
Marc Demange, RMIT University, School of Science - Mathematical Sciences, Australia
A Challenge for an Efficient Supply Chain Management - Integration Versus Optimality
Marino Widmer, DIUF, University of Fribourg, Switzerland
Analyzing the Performance of Greedy Maximal Scheduling via Local Pooling and Graph Theory
Bernard Ries, Université Paris-Dauphine, France
Hybridizing Multi-Objective Interactive Techniques - Some Applications to the Electricity Generation Industry
Francisco Ruiz
University of Málaga
Spain
![]()
Brief Bio
Francisco Ruiz is a full professor of Quantitative Methods for Economy at the University of Málaga in Spain. He teaches at graduate courses in Economy, Business Administration and Marketing, in the Masters Course in Quantitative Methods for Economy, and in the PhD program in Economy and Business Administration. His research activities are focused on multi-objective programming methodologies, mainly in interactive methods and their applications to decision making processes in industry, development of synthetic indicators and portfolio selection. Presently, he is the Secretary of the International Society on Multiple Criteria Decision Making (www.mcdmsociety.org).
Francisco Ruiz received his BSc degree in Mathematics in 1989, and his PhD degree in Economy and Business Administration in 1994, both at the University of Málaga. He has carried out his professional activity at this institution, where he has held different positions until he became a full professor in 2009. He has stayed as visiting researcher at the University of Georgia (USA) and at the University of Manchester (UK).
Francisco Ruiz is the (co)author of more than 50 publications (including papers in professional journals, proceedings of conferences, and book chapters) and has reviewed papers for more than 20 international journals. He has participated in more than 30 research projects, and he has had research activities with several Universities in Europe and North America, as well as with public and private companies.
Abstract
Interactive methods have proved to be extremely useful multi-objective techniques, when it comes to solve real complex decision making problems. Their iterative schemes are especially suitable for the necessary learning process that has to be present in very decision making process. Many different interactive methods exist, and they vary both in the type of information that the decision maker (DM) has to provide at each iteration, and in the way the different solutions are obtained along the process. These two aspects have leaded us to develop several hybrid models:
- With respect to the information required from the DM, this can take many different forms (just choosing one solution among a set of possible solutions, giving local tradeoffs, giving reference or target values, classifying the objectives…). But in many cases, the interactive method is chosen without taking into account the cognitive burden that it implies for the DM. In this sense, we have developed a hybrid interactive multi-objective system (PROMOIN), where the DM can decide at each step the type of information (s)he prefers to give, and the system internally switches to the most appropriated method. The idea is to adapt the resolution process to the necessities of the DM, and not vice versa. PROMOIN was applied to an empirical study, founded by the Regional Ministry of Environment of Andalucía, about the determination of the optimal electricity mix, taking into account economical and environmental criteria.
- With respect to the inner methodology, interactive methods imply the sequential resolution of single objective optimization problems, which, depending on the specific features of the model, can be very complex and time consuming. In many of these cases, traditional solvers are no longer applicable. For these reason, we have studied several hybridizations of interactive methods with Evolutionary Multi-Objective techniques, which have proved to be very useful, not only to solve the individual optimization problems, but to extract many useful information during a preliminary learning phase. We have applied these hybrid methodologies to two real decision problems of one of the main Spanish electricity generation firms. In one of them, we studied the optimal dimensions of a solar thermal generation plant. In another one, we aimed in the decision of the actions to take in a coal thermal plant, in order to improve the efficiency of the auxiliary systems.
Online Strategies for Hard Optimization Problems in Graphs
Marc Demange
RMIT University, School of Science - Mathematical Sciences
Australia
Brief Bio
Marc Demange holds a PhD in Computer Science from Paris I - Pantheon Sor-
bonne University (1994) and an Habilitation Thesis in Computer Science from
Paris Dauphine University (2000). He has held the positions of Assistant Profes-
sor in Computer Science at Paris 1 Panthon Sorbonne University, Associate Pro-
fessor and Professor in Operational Research at ESSEC Business School (Paris -
Singapore), where he has been Vice Dean of the Faculty, Associate Dean for Re-
search and Director of ESSEC Romania Centre (in Bucharest). Since 2014 he is
Associate Professor in Mathematical Sciences at RMIT University, Melbourne,
Australia, presently in charge of the Bachelor in Mathematics. During his career
he has taught a large range of topics in Computer Science, Operational Research
and Discrete Mathematics. His research interests, in Combinatorial Optimisa-
tion, are centred around the notion of ecient algorithms with performance
guarantees, mainly polynomial approximation, complexity theory, algorithmic
graph theory, online algorithms and inverse combinatorial optimisation. He is
(co)author of more than 50 papers in international journals or book chapters.
Abstract
Online optimization deals with instances of problems that are incrementally revealed over time. One should take decisions as soon as data are presented without information about the future data. The objective is to design strategies with absolute performance guaranties on the quality of the online computed solution vs the best possible solution with complete information. Online optimization allows to address dynamic environments with continuous flow of data, in particular when it is hard to assign a relevant probability system on possible futures. This approach reveals to be particularly interesting nowadays as more and more applications request to take into account ongoing instantaneous data.
From a theoretical perspective, the online framework allows to better understand how the solutions of a problem may be a affected by a lack of information. In some sense, this topic can be seen as a continuation of approximation theory. Many concepts of polynomial approximation, such as reductions, could still be transferred to the online case.
We will discuss different examples in online graph optimization (notably online coloring and TSP) that illustrate the basis of online algorithms and emphasize some future plans of this topic.
A Challenge for an Efficient Supply Chain Management - Integration Versus Optimality
Marino Widmer
DIUF, University of Fribourg
Switzerland
![]()
Brief Bio
Marino Widmer is full professor at the Department of Informatics - University of Fribourg (Switzerland). He currently offers courses at the bachelor level (mathematics, operations management) as well as at the master level (supply chain management and logistics). In order to propagate operations research concepts in small and medium enterprises, all his students’ theses are based on real life projects, in close cooperation with companies.
He organized several international conferences (such as seven “GO meetings” (on graphs and optimization) and one FRANCORO, in 2004) and was also active in 47 program committees. In March 2000, he organized with Alain Hertz (Ecole Polytechnique de Montréal) the XVIIIth EURO Winter Institute, dedicated to the research theme “Metaheuristics in Combinatorial Optimisation”. This event was at the origin of one of the most active EURO working groups: EU/ME: the metaheuristics community (www.metaheuristics.eu).
His major research interest deals with the development of quantitative models and methods of operations research with their applications in manufacturing and logistics (design and layout; production planning; scheduling; real time control, delivery). The activities include theoretical and methodological research at the level of models, algorithms and systems. His current research efforts are in combinatorial optimization (in particular scheduling theory), heuristic and metaheuristic methods, simulation methods and decision support systems. His research is typically conducted in connection with targeted applied problems where novel models, methods and systems can generate substantial added value. Such opportunities occur at the design stage as well as at the operational stage of complex systems. The application areas include supply chain and distribution management, production planning and control in industry and services, design and control of systems in manufacturing and logistics.
He is currently permanent treasurer of EURO (www.euro-online.org), the “Association of European Operational Research Societies“ within IFORS, the “International Federation of Operational Research Societies“ (ifors.org/web/).
Abstract
According to Hartmut Stadtler[1], a supply chain is a “network of organizations that are involved, through upstream and downstream linkages, in the different processes and activities that produce value in the form of products and services in the hands of the ultimate customer“. Moreover, he defines the supply chain management “as the task of integrating organizational units along a supply chain and coordinating material, information and financial flows in order to fulfill (ultimate) customer demands with the aim of improving competitiveness of a supply chain as a whole“.
These definitions highlight the importance of integration and coordination in an efficient supply chain management. However, if one analyzes the models detailed in the scientific literature, none of them proposes to solve the problem of supply chain management in its entirety. In most cases, models are formulated to solve optimally a sub-problem (inventory management, delivery tour, ...). In rare cases, models link together two or in the best cases three components of the supply chain (for example: “production – delivery” or “production – warehousing – delivery”). This is mainly due to the fact that the development of a global model is practically impossible (too many constraints and variables).
In the current state of the scientific knowledge and the technology, solving the global problem of the supply chain management remains a utopia. However, it is possible to be more efficient in the resolution of complex problems through the use of new approaches, such as matheuristics (optimization algorithms combining the efficiency of metaheuristic methods with the accuracy of mathematical programming techniques) or kernel search (a general heuristic approach to mixed-integer linear programming (MILP) problems).
The purpose of this presentation is mainly to describe a general framework for the supply chain management, in order to identify each sub-problem in its context, and to illustrate the application of new tools (such as matheuristics and kernel search) in solving several sub-problems related to supply chain management.
[1] in Hartmut Stadtler & Christoph Kilger (Eds), “Supply Chain Management and Advanced Planning“, 4th edition, Springer (2007)
Analyzing the Performance of Greedy Maximal Scheduling via Local Pooling and Graph Theory
Bernard Ries
Université Paris-Dauphine
France
Brief Bio
Bernard Ries obtained his M.Sc. in 2004 at the Ecole Polytechnique Fédéral de Lausanne (Switzerland). He finished his PhD thesis in 2007 at the same institute. In 2008-2009, he was a Postdoctoral Research Fellow at Columbia University (USA) in the Department of Industrial Engineering and Operations Research. In 2009-2010, he was an Assistant Professor in the Centre for Discrete Mathematics and its Applications (DIMAP) at the University of Warwick (UK). He was also member of the Operational Research and Management Sciences group (ORMS) at Warwick Business School. Since September 2010, he is an Associate Professor at the Université Paris Dauphine in the research group LAMSADE. He has been invited researcher at several Universities: CNAM (France), Memorial University (Canada), University of Haifa (Israel), University of Warsaw (Poland), Université de Montréal (Canada), Durham University (UK), Gdansk University (Poland), University of Luxembourg (Luxembourg). His research interests are structural and algorithmic graph theory, combinatorial optimization and computational complexity. He has published more than 40 papers in international journals and conference proceedings. In 2004, he was given the SORS Master Thesis Award offered by the Swiss OR Society for best master thesis in OR in Switzerland and from 2014 to 2017, he is a holder of the "Prime d'Excellence Scientifique" (bonus for scientific excellence).
Abstract
Efficient operation of wireless networks and switches requires using simple (and in some cases distributed) scheduling algorithms. In general, simple greedy algorithms (known as Greedy Maximal Scheduling - GMS) are guaranteed to achieve only a fraction of the maximum possible throughput. However, it was recently shown that in networks in which some specific conditions (called the Local Pooling conditions) are satisfied, GMS achieves 100% throughput. Therefore, it is of interst to know exactly which networks satisfy these conditions, i.e. to identify the specific network topologies that satisfy the Local Pooling conditions. Using structural graph theory, we provide the first characterization of all the network graphs in which Local Pooling conditions hold under primary interference constraints (in these networks GMS achieves 100% throughput). This leads to a linear time algorithm for identifying Local Pooling-satisfying graphs.
This is joint work with Berk Birand, Maria Chudnovsky, Paul Seymour, Gil Zussman and Yori Zwols.