Seminar

Robust risk-aware dynamic programming

19 May 2022
Start time 
4:30 pm
PovoZero - Via Sommarive 14, Povo (Trento)
Sala Seminari "-1"
Target audience: 
University community
Attendance: 
Online – Registration required
Registration email: 

“Doc in Progress” and #iorestoacasa are pleased to introduce you to:

  • Marta Fochesato - PhD in Information Technology and Electrical Engineering, ETH Zürich

Abstract:
Stochastic optimal control has numerous applications in engineering and science. It refers to a sub-field of control theory that deals with the
existence of uncertainties that drive the evolution of a system in an unpredictable manner, posing several issues in terms of computational tractability and robustness. On one side, the computational issue is dealt with in the literature relying on parametric value function approximations. On the other side, the
robustness has been recently addressed borrowing tools from the field of distributionally robust optimization.
In this context, armed with tools from optimal transport theory, we propose a new computationally efficient dynamic programming paradigm, showing improved out-of-samples performance. Additionally, we extend the framework so as to handle probabilistic constraints in the state, proposing a primal-dual risk-aware formulation.

The seminar will be held both in presence in Seminar Room "-1" (Povo 0) and online via Zoom.

To join the event, please contact docinprogress.unitn [at] gmail.com using an institutional e-mail address for both reserving a sit in the seminar room or obtaining login credentials.