Stochastic Optimization with Multiple Time Scales

Real-world multistage stochastic optimization problems are often characterized by the fact that the decision maker may take actions only at specific points in time, even if relevant data can be observed much more frequently. In such a case there are not only multiple decision stages present but also several observation periods between consecutive decisions, where profits/costs occur contingent on the stochastic evolution of some uncertainty factors. We present a tailor-made modeling framework for such problems, which allows for a computationally efficient solution. We first establish new results related to the approximation of (Markovian) stochastic processes by scenario lattices. In a second step, we incorporate the multiscale feature by leveraging the theory of stochastic bridge processes. The ingredients to our proposed modeling framework are elaborated explicitly for various popular examples, including both diffusion and jump models. In particular, we present new results related to the simulation of compound Poisson bridges. Finally, we discuss a valuation problem of a thermal power plant, where implementing our multiscale modeling framework turned out to be particularly convenient. If time permits, we incorporate model ambiguity into the power plant valuation problem and show some numerical results.