This workshop targets the niche area of research activity at the intersection of learning theory and control.

Control theory is a thoroughly studied subject with an extensive technical base that has evolved over more than a century in its modern avatar. Its foundational principles relied on the availability of physics-driven models, and a vast array of technical machinery have been established for the purpose of explaining the behavior of dynamical systems on the descriptive side, and for controlling such dynamical systems on the prescriptive side. The emphasis of the subject has historically focused more on the front of direct problems.

Learning is a central topic of today’s research activities across the globe, and has impacted a bewildering array of technical disciplines over the recent past. Given the enormity of the raw data and features, modern learning theory focuses on scalability issues. In particular it is important to understand the learning algorithms that can extract (latent) structures in large scale problems and adapt to it. Tremendous advances in this subject have been made over the past few decades, thanks to vast storage capacities and raw computational power.

These two pillars of the broad area of cybernetics are evolving rapidly today. This workshop targets the substrate common to both these disciplines and an exposition of the chief technical tools that drive both, and promises to provide a platform for brisk discussions and exchange of ideas.

Student Coordinators

• Aman Batra, IIT Bombay, India (aman@sc.iitb.ac.in, 214230006@iitb.ac.in)

• Ashwin Aravind, IIT Bombay, India (ashwin@sc.iitb.ac.in, a.aravind@iitb.ac.in)