The invited speakers who have confirmed their interest in participating are:
- Rosiane de Freitas
Professor, Universidad Federal de Amazonas, Brasil.
Topic: Combinatorial Optimization and Computational Biology
- Mario Guajardo
Assistant Professor, NHH Norwegian School of Economics, Norway.
Topic: Sports Analytics
- Denis Sauré
Assistant Professor, Universidad de Chile, Chile.
Topic: Sports Analytics
- José Mario Martínez
Professor, IMECC Universidade Estadual de Campinas, Brasil.
Topic: Applied Nonlinear Programing
- Ana Friedlander
Professor, IMECC Universidade Estadual de Campinas, Brasil.
Topic: Applied Nonlinear Programing
- Luciana Buriol
Professor, Universidad Federal de Rio Grande do Sul, Brasil.
Topic: Integer Programming and Network Optimization
- Martín Safe
Professor, Universidad Nacional del Sur, Argentina.
Topic: Intersection Graphs and Perfect Graphs
- Luciano Grippo
Professor, Universidad Nacional de General Sarmiento, Argentina.
Topic: Intersection Graphs and Perfect Graphs
- Cristián Cortés
Professor, Universidad de Chile, Chile.
Topic: Transportation
- Pablo Rey
Professor, Universidad de Chile, Chile.
Topic: Transportation
- Stefano Smriglio
Professor, University of L'Aquila, Italy.
Topic: Combinatorial Optimization
Titles and abstracts
Talks (February 24 th., University of Buenos Aires)
- Javier Marenco (UBA and UNGS, Argentina)
Title: A review of applied operations research projects at the Instituto de Cálculo (2007-2016)Abstract: Operations research covers a wide range of topics, and can be applied in many environments. In this talk we will review projects carried out at the Instituto de Cálculo (FCEN-UBA, Argentina) during the last decade, including applications to logistics, auctions, and sports. We shall discuss features and particular issues of projects with private companies, public institutions, and sports organizations, and we will point out some lessons that we have learned along the way.
- Guillermo Durán (UBA, Argentina and U. de Chile, Chile)
Title: Operations Research transforms the scheduling of Chilean soccer leagues, the South American World Cup Qualifiers and other sports leagues in Argentina
Abstract: For the past 12 years, the Chilean Professional Soccer Association (ANFP) has applied operations research (OR) techniques to schedule soccer leagues in Chile. We have scheduled more than 50 tournaments using this approach, resulting in an important economic impact that includes reductions in television broadcaster operating costs, growth in soccer pay-television subscriptions, increased ticket revenue, and lower travel costs for the teams. This application of OR has also had significant noneconomic impacts. First, the incorporation of team requirements and various sporting criteria has improved process transparency and schedule fairness, increasing fans’ interest in local professional tournaments; second, because of the high portability of these techniques, they have been used successfully to schedule sports leagues in other countries; examples include volleyball and basketball in Argentina, and the South American qualifiers for the 2018 Soccer World Cup. Furthermore, the models and methods used in this scheduling application have been disseminated widely, helping to promote OR as an effective tool for addressing practical problems. This project was selected as one of the six finalists in the 2016 Franz Edelman Award competition, organized by INFORMS. This competition recognizes outstanding examples of real-world advanced analytics and operations research projects that are transforming industries, companies, and people’s lives.
-
Stefano Smriglio (University of L'Aquila, Italy)
Title: Strong relaxations and cutting planes for the stable set problemAbstract: A great deal of research has been focusing, since the early seventies, on finding strong relaxations for the stable set problem. Polyhedral combinatorics techniques have been at first developed to strengthen the natural linear formulation. Afterward, strong semidefinite programming relaxations have been deeply investigated, starting from the celebrated Lovasz theta function. Nevertheless, the resulting integer programming (IP) algorithms cannot be regarded as being completely successful in practice, as most of the relaxations give rise to one out of two extreme situations: either provide weak bounds at low computational cost or give good bounds (sometimes excellent) but too demanding to compute. In this talk we present two of our attempts to bridge such a gap. The first attempt is based on a convex programming relaxation whose feasible region takes the form of an ellipsoid. The nice property is that the optimal value of this relaxation is equal to the Lovasz theta number, even though the
ellipsoid has a more friendly structure than the semidefinite relaxation. The second direction consists of looking at the classical Chvatal-Gomory cuts and strengthening their right-hand sides. Although the strengthening problem is itself NP-hard, it can be solved reasonably quickly in practice. We show that both the methods yield competitive upper bounds in reasonable computing times. - Flavia Bonomo (UBA, Argentina)
Title: Algorithms for polynomial instances of graph coloringAbstract: In this talk we will review some of the graph classes in which the graph coloring problem can be solved in polynomial time (chordal graphs, proper circular-arc graphs, graphs with no induced paths of certain lengths, etc.), as well as the algorithms themselves.
Courses (From February 27th. to March 3rd., Miramar)
- Ana Friedlander (Unicamp, Brazil) and José Mario Martínez (Unicamp, Brazil)
Title: Convergence and complexity in Nonlinear Optimization
Abstract: Methods for minimization of continuous functions; results on asymptotic convergence; results on Local Convergence; meaning of complexity in Nonlinear Optimization; complexity of gradient-like and Newton-like methods; challenges in complexity for Constrained Optimization.
- Mario Guajardo (NHH, Norway) and Denis Sauré (U. de Chile, Chile)
Title: Applications of Operations Research and Statistics to Sports Analytics
Abstract: In this course, we study the use of methods from operations research, statistics and economics to sport analytics. In a first part, we study the use of classic operations research methods to scheduling tournaments. In particular, we will survey the use of OR techniques to model and solve tournament scheduling for professional sports. Our exposition here focuses on practical implementation and resolution techniques, and is based in our experience scheduling professional soccer leagues in Chile and South America. On a second part, we review the use of methods from statistics and economics for predicting outcomes of games in professional sports. Then, we study in more detail the use of analytics for predicting outcomes of tournaments, and its use for developing successful gambling strategies in the context of fantasy leagues competitions.
- Cristián Cortés (U. de Chile, Chile) and Pablo Rey (U. de Chile, Chile)
Title: Integer programming models for planning and operation of public transit systemsAbstract: In any modern large city, public transport systems are among the most necessary and relevant services to make it work. Efficient planning and operation of such systems are problems that have been studied for years. For some of the key stages related to transit planning, various techniques have been developed based on optimization models. The use of such models have allowed efficient design, planning and operation of massive transport systems in big cities during the last decades. Lately, the new technologies and information systems available bring new challenges to expand and improve the quality of service for users of such systems. In this course, we will first introduce concepts and basic problems in public transport planning and discuss some basic models used in this context. In a second part, we present some challenges in modeling we are currently facing in the context of the development of the public transport system of the city of Santiago (Chile).
- Luciano Grippo (UNGS, Argentina) and Martín Safe (UNS, Argentina)
Title: Intersection graphs and perfect graphs
Abstract: In this course, we will present some special graph classes, their structural properties, and algorithmic problems related to them. The introduction of these special graph classes is justified because they arise naturally in connection to certain optimization problems, either as models or as tractable cases for problems which are considered intractable in the general case. The course will be divided into two parts. In the first part, we will present some intersection graph classes. In the second part, we will discuss about the class of perfect graphs and some of its variants. The presentation is divided into two parts for ease of exposition but these two parts do not exclude each other and, in fact, several interesting interactions between them will be highlighted along the course.
- Luciana Buriol (UFRGS, Brazil)
Title: Mathematical Formulations for Integer Programming Problems
Abstract: This tutorial aims at presenting and discussing IP formulations for several combinatorial optimization problems. Initially we discuss how to model basic restrictions, and then we formulate several problems. During the tutorial we will formulate a few graph problems (independent set, graph coloring, point-to-point shortest path, network flow), traveling salesman problem, vehicle routing problem (capacitated and time windows), high school time tabling, among other problems.
- Rosiane de Freitas (UFAM, Brazil)
Title: Molecular Distance Geometry Problem: concepts, algorithms and computacional complexity
Abstract: In this tutorial we discuss about computational and mathematical theoretical concepts and algorithms for determining the 3D structure of a protein molecule, represented by a simple graph with some arbitrary weighted edges to represent a set of atoms and a subset of their distances in a molecule, The Molecular Distance Geometry Problem (MDGP) is an NP-hard computational problem for an incomplete and arbitrary set of distance constraints, and a Polynomial problem when all the distance set is known. Continuous and discrete mathematical approaches to solve MDGP are revised, based on the analysis of two types of calculating sphere intersection: solving systems from interatomic Euclidean distance equations, or solving internal coordinate systems using matrix multiplication techniques.