The course covers the fundamentals of dynamic model learning, covering experiment design, classical and contemporary model classes such as, e.g., ARMAX and Gaussian Process models, as well as state-of-the art estimation algorithms, building on powerful numerical algorithms such as Markov Chain Monte Carlo methods, the EM-algorithm, SVD, convex relaxations, and multi-step least-squares.

 

The course is intended for PhD students in all areas that come in contact with data-driven modeling of dynamical systems. The lectures are aimed at providing the, in fact, few underlying fundamentals and ideas that underpin this area so that participants can make informed choices on how to tackle specific applications, while details are to be obtained from reading material. The underlying theory as well as practical exercises are included.  

An extended version of this course is FEL3202 Data Driven Modeling - Extended Course (12 credits), intended for PhD students working on research topics in or related to this area. The two courses are given simultaneously, with FEL3202 including more lectures on theory and homework problems, as well as a more extensive exam.