Decomposition Methods for Multi-stage Stochastic Linear problems

Specialeforsvar ved Sara el-Wali

Titel: Decomposition Methods for Multi-stage Stochastic Linear problems

Resume: Many operational and planning problems involve sequences of decisions over time, which are contingent upon realizations of outcomes that are not known a priori.These problems are thus usually formulated as multi-stage stochastic linear programming problems. However, the complexity of these problems increases exponentially with the number of stages. The aim of this project is to explain and compare two different decomposition methods to reduce the computational burden of multi-stage stochastic programming problems 

Vejleder:  Salvador Pineda Morente  
Censor:    Niklas Kohl