Data-driven methods for optimal control

In this short course, we will study how to incorporate elements of machine learning into optimal control design. The course is split into 3 parts:
    1    Fundamentals of optimal control: dynamic optimization, linear-quadratic control, dynamic programming and Pontryagin's maximum principle. Nonlinear optimal control.
    2    Approximation methods in high dimensions: polynomial approximation, deep neural networks. Optimization techniques: LASSO regression, stochastic gradient descent, training neural networks.
    3    Synthetic data-driven schemes for optimal control. Combining the first two parts, we will study the construction of data-driven schemes for the approximation of high-dimensional nonlinear control laws. Examples in Matlab.

The schedule of the course is:

• Friday     02/12 9:00.  - 10:30, room C
• Monday  05/12 9:00.  - 10:30, room G
• Tuesday 06/12 10:30 - 12:30 Sala riunioni secondo piano

Students interested are invited to contact Prof. Albi (giacomo.albi@univr.it)