Lab Home | Phone | Search | ||||||||
|
||||||||
Dynamic Mode Decomposition (DMD) is a methodology through which a model for system state corresponding to unknown dynamics may be extracted from observations. The execution of DMD involves certain operators posed over Hilbert functions spaces. In this talk, we will outline the general theory and goals of Kernel Based Dynamic Mode Decompositions. We will cover vector valued reproducing kernel Hilbert spaces, convergence within Hilbert spaces, and dynamic operators. This will reveal some inherent challenges in DMD, and we will conclude by discussing how they might be mitigated. Host: Nick Hengartner |