Thaleia Zariphopoulou: Learning and Stochastic Optimization

Submitted by Tony I Garcia on

The Department of Applied Mathematics is pleased to host this series of colloquium lectures, funded in part by a generous gift from the Boeing Company. This series will bring to campus prominent applied mathematicians from around the world.

Speaker: Thaleia Zariphopoulou

Date: April 12th, 2018, 4pm, reception to follow

Location: (SMI 102

Title: Learning and Stochastic Optimization

Abstract: Classical stochastic optimization models require an a spriori choice of a criterion, a horizon and a model (or a family of models) for the dynamics of the controlled and uncontrolled processes. The solution and the optimal policies are then specified backwards in time, through the dynamic programming principle. This setting, however, precludes the incorporation of realistic “forward-in-time” elements, like learning, model and preferences revision, rolling horizons, and others. In this talk, I will introduce a new approach, the forward stochastic optimization, and focus on how learning can be incorporated in model revision, optimality and time-consistency. I will discuss the associated stochastic partial differential equation, and present applications in asset allocation and resource management, with and without competition among agents.