Data-Driven Control
Complex dynamical systems usually have to collaborate computational elements controlling physical entities. The composition of continuous and discrete models is essential for capturing the behavior of such systems. Verification and synthesis of these dynamical systems are algorithmically studied using abstraction techniques and model checking tools. In this research, towards the goal of developing principled data-driven methods for control and decision-making, we study the interplay between formal verification, machine learning, and dynamics in real-world systems. Application areas of the research include, among others, the efficiency of buildings and systems biology.
Data-Driven Control
Complex dynamical systems usually have to collaborate computational elements controlling physical entities. The composition of continuous and discrete models is essential for capturing the behavior of such systems. Verification and synthesis of these dynamical systems are algorithmically studied using abstraction techniques and model checking tools. In this research, towards the goal of developing principled data-driven methods for control and decision-making, we study the interplay between formal verification, machine learning, and dynamics in real-world systems. Application areas of the research include, among others, the efficiency of buildings and systems biology.