When working with stochastic programming problems, we frequently encounter optimization problems that are too large to be processed by routine methods of mathematical programming. However, in some cases the problem structure allows for a use of specialized decomposition methods that (when utilizing said structure) can be employed to efficiently solve very large optimization problems. This work focuses on two classes of  stochastic programming problems that have an exploitable structure,
namely two-stage stochastic programming problems and chance constrained problems, and the advanced decomposition methods that can be used to solve optimization problems in these two classes. We describe a novel warm-start cuts for the Generalized Benders Decomposition, which is used as a methods for the two-stage stochastic programming problems. For the class of chance constraint problems, we introduce an original  decomposition method, that we named the Pool & Discard algorithm. The usefulness of the described decomposition methods is demonstrated
on several examples and engineering applications.