Stochastic optimal control on positive cones

Title: Stochastic optimal control on positive cones
Speaker:  Mohamad Al Ahdab, Postdoc at Aalborg University, Denmark
Date & Time: June 23rd, 13:15-14:00
Location: Seminar Room M 3170-73 in the M-building, LTH

Abstract: 
In this presentation, I will formulate and study stochastic optimal control problems on proper cones under risk-neutral and risk-sensitive average-cost criteria. I will introduce verifiable model 
classes for which the Bellman recursion admits an explicit finite-dimensional representation, and then focus on the polyhedral cone case as a detailed example. Finally, I will briefly discuss how suitable large-risk scalings of state-dependent disturbances connect the risk-sensitive formulation to minimax control of positive systems.

Bio: 
Mohamad Al Ahdab is a postdoctoral researcher at Aalborg university. He is currently working on optimal control problems, with a particular interest in adaptive and dual control. Mohamad has a PhD from Aalborg university, where he worked in collaboration with Novo Nordisk on modeling, estimation, and feedback strategies for type 2 diabetes and contributed to implementing decision support algorithms for a clinical evaluation.