Optics corrections using machine learning in the LHC

Optics corrections in the LHC are based on a response matrix between available correctors and observables. Supervised learning has been applied to optics correction in the LHC demonstrating promising results on simulations and demonstrating the ability to reach acceptably low $\beta$-beating. A comparison of different algorithms to the traditional response matrix approach is given, and it is followed by the presentation of further possible concepts to obtain optics corrections using machine learning (ML).