Stochastic programming
3-4 hours of lectures per week.
Lecturer: Morten Riis
Content: Stochastic programming is concerned with optimization problems in which some parameters are replaced by random variables in order to capture the uncertainty which is almost always an inherent feature of the system being modelled. This course focuses on the class of stochastic recourse programs in which an alternating process of decisions and observations of random outcomes in a finite number of stages is assumed. Structural properties and solution procedures for two-stage stochastic linear recourse programs will be discussed in detail and extensions such as integer recourse models and multistage problems will be considered.
Prerequisites: Convex analysis and Mathematical programming.
Assessment: The basis for passing the course is active participation at the lectures. Also, satisfactory answers to a number of assignments must be produced.
Literature: J.R. Birge and F. Louveaux: Introduction to Stochastic Programming. Springer-Verlag, 1997. Furthermore, a few articles will be used.