PhD course: Introduction to Stochastic Programming

The course has been cancelled due to the coronavirus situation

Date: 20 - 24 April 2020

Location: Aarhus BSS, Aarhus University, Fuglesangs Allé 4, 8210 Aarhus V, Denmark

LecturerProfessor Walter Rei, Department of Management and Technology, University of Quebec in Montreal (UQAM)

Host: Cluster for Operations Research, Analytics, and Logistics (CORAL)


In organizational decision making, the vast majority of planning and operational decisions are made without a complete knowledge of the current situation and the future. Plans that are made when part of the contextual information is uncertain are then adjusted when uncertain information is revealed, thus allowing for recourse actions to be taken. In all generality, this leads to problems where sequences of decisions are made over varying time horizons and where the overall aim is to balance both the immediate benefits of the taken decisions and their impact in the future. For such problems, designing optimization methodologies that 1) explicitly consider the various sources of uncertainty that are present and 2) efficiently produce high-quality solutions for these problem settings, is quite challenging

The general objectives of the present course are:

  1. Introduce students to the overall approach that is used to formulate an optimization problem that involves uncertainty as a stochastic program. This approach entails, identifying the specific problems for which stochastic programming is an appropriate method to apply, modelling feasibility and the dynamics of the problem and formulating the objective function.
  2. Present the main specialized solution methods that have been developed to solve stochastic programs. Specifically, a review of the principal exact solution methods and heuristic algorithms will be provided. The focus here will be to present how more efficient solution processes can be obtained by applying a series of tailored strategies, i.e., scenario generation, mathematical decomposition, heuristic search principles, etc.