Modern internetworked systems are characterized by dynamicity and stochasticity. It is of high research and practical importance to investigate and study how the network and its structure affect the dynamic properties of the overall system, including stability, performance, and robustness. Considering both random and deterministic settings, different decision making problems arise dealing with both the resource sharing in these systems, as well as with various system-theoretic aspects such as controllability and observability. Our research focuses on the combination of both model-driven approaches and artificial intelligence based approaches in order to predict, characterize and control the behavior of such networked systems. In particular we use a set of mathematical modeling tools from game theory, prospect theory, control theory and machine learning, while also dealing with real-world problems such as the incomplete/partial information availability, environment dynamicity and stochasticity.