Antrittsvorlesung von Max-Planck-Fellow Prof. Dr. Sebastian Sager: Generalized Inverse Optimal Control
- Datum: 23.11.2023
- Uhrzeit: 14:00 - 16:00
- Vortragender: Prof. Dr. Sebastian Sager, Max-Planck-Fellow
- Prof. Dr. Sebastian Sager, Forschungsgruppe „Mathematical Optimization and Machine Learning (OML)“
- Ort: Max Planck Institute Magdeburg
- Raum: Großer Seminarraum "Prigogine"
- Gastgeber: Max Planck Institute Magdeburg
- Kontakt: email@example.com
We survey the foundations of generalized inverse optimal control (gIOC), a mathematical approach to infer optimality principles from dynamic data. With this technique it is possible to elucidate and forecast dynamics in mathematical biology, with a particular focus on molecular and cell biology. Our motivation stems from the observation that numerous biological systems seem to be optimized due to evolution. However, the specific objectives, and possibly also constraints and dynamics, are (partially) unknown a priori. gIOC aims to infer them from data and partial prior knowledge, which can then be used to understand the system's behavior and to design optimal intervention strategies. Handling gIOC with nonlinear dynamics is a formidable mathematical challenge: even small-scale instances are currently not solvable, due to a nested bi-level optimization structure on top of noisy data and nonlinear dynamics. Combinatorial structures arise on several layers, e.g., via complementarity constraints of vanishing constraints, which have not yet been investigated in the literature.
Im Anschluss an den Vortrag lädt Sebastian Sager zu einem
kleinen Buffet ein. Bitte registrieren Sie sich bis zum 17. November zur
besseren Planbarkeit unter: