Modern algorithms for learning, optimization, and control operate in dynamic environments, interacting with physical systems, data streams, and other algorithms. This modern landscape challenges the traditional view of algorithms as isolated "pieces of code", and motivates interpreting them instead as open dynamical systems evolving in feedback with their environment.

This workshop explores the application of systems theory and control tools for the analysis and design of modern computational strategies in learning and optimization. The discussion will focus on two main themes, connected via the systems theory perspective:

 1) The analysis and design of algorithms using systems theory,

 2) Real-time algorithms in feedback loop with the world.

We will also outline open problems and fresh future research directions in these two domains using tools from systems theory.

Building on the success of the first edition of this workshop - the largest workshop at ECC 2025 - this year’s program brings together a new set of experts from leading institutions, including MIT, ETH Zurich, Cambridge, Boston University, KTH, TU Eindhoven, and others.

Speakers