This document presents the final report on the research investigations and outcomes of our AFOSR DDDAS project titled Stochastic Hybrid Systems Modeling and Middleware-enabled DDDAS for Next-generation US Air Force Systems. It summarizes our contributions to the various facets of the DDDAS paradigm when applied to provide dynamic resource management in cloud computing platforms so that they can support applications with different quality of service requirements. To that end, first, we describe our approach on workload characterization of cloud-hosted applications using online model learning that is used for resource management in the cloud. Second, we report on a new service called Simulationbased Optimization as a Service, which is an approach we have developed to simulate the learned models to obtain optimal values of parameters to a model that are applied to the system in the DDDAS feedback loop. Third, we report on a number of dynamic resource management algorithms we have developed and their experimental evaluations for hosting DDDAS-like applications in public cloud infrastructures. Finally, we report on ongoing work towards using the DDDAS paradigm in the continuum from cloud to the edge to support applications that are hosted across the cloud-edge spectrum. Keywords-Dynamic resource management, model learning, simulation-based optimizations, cloud infrastructures for DDDAS applications.