Study Programmes 2015-2016
GDOC0005-1  
Seminar of Operations Research
Duration :
30h Th
Number of credits :
Doctoral training in economics and business management (Sciences de gestion)5
Lecturer :
Yves Crama
Language(s) of instruction :
English language
Organisation and examination :
Teaching in the second semester
Units courses prerequisite and corequisite :
Prerequisite or corequisite units are presented within each program
Course contents :
The main focus of the seminar is on optimization methods, at an advanced level. The main topics addressed are:


  • complexity theory: easy and hard optimization problems, complexity classes, polynomial transformations
  • linear programming theory: (revised) simplex algorithm, duality theory, column generation algorithms
  • combinatorial algorithms: network flows, shortest paths, spanning trees, etc.
  • integer progamming: branch-and-bound, tight formulations, introduction to cutting planes and to polyhedral theory
  • approximation algorithms with guaranteed performance.
The selection of topics can be adapted to a certain extent, depending for instance on the research interests of the participants.
Learning outcomes of the course :
The course is mostly intended for doctoral students pursuing methodological advances in operations research and in optimization, or applying OR approaches to the solution of problems arising in disiciplines such as supply chain management, transportation, marketing, finance, industrial economics, and so on.
Intended learning outcomes
  • Knowledge: acquisition and in-depth understanding of linear and discrete optimization (theory and algorithms)
  • Specific skills: ability to formulate and to solve optimization problems arising in management
  • Specific skills: ability to use methodological knowledge in order to develop new algorithms and solution approaches
  • Specific/transversal skills: ability to read and to understand the scientific literature in these fields
  • Transversal skills: improvement of oral presentation skills
Prerequisite knowledge and skills :
Prerequisites:
  • A first course in operations research, including linear programming models and the simplex method.
  • Proficiency in mathematics, especially linear algebra and analysis.
  • Introduction to algorithms.
Planned learning activities and teaching methods :
The students must prepare each meting by reading preassigned material (either chapters from advanced textbooks or research articles) and by solving homework problems. The homework problems are intended to clarify difficult concepts, as well as to deepen and to test the understanding of the material.
Classroom meetings are devoted to group discussions of the material and of the homework assignments.
A few meetings may be devoted to individual presentations by the students, based on their general scientific interests or on their own research projects.
Mode of delivery (face-to-face ; distance-learning) :
Group discussions and presentations by the students
Recommended or required readings :
Required readings: chapters from several books, mostly
Bertsimas and Tsitsiklis, Introduction to Linear Optimization, Dynamic Ideas and Athena Scientific, Belmont, Massachusetts, 2008.
Chvátal, Linear Programming, WH Freeman & Co, San Francisco, 1983.
Cook, Cunningham, Pulleyblank and Schrijver, Combinatorial Optimization, John Wiley and Sons, New York, 1998.
Wolsey, Integer Programming, Wiley & Sons, New York, 1998.
Assessment methods and criteria :
The final grade is based on:
  • homework assignments 60%;
  • student's involvement (presence, density and quality of participation and presentations) 40%.
Work placement(s) :
Organizational remarks :
Contacts :
Prof. Yves Crama
Y.Crama@ulg.ac.be