engleski

Parametrizing Compton form factors with neural networks

We describe a method, based on neural networks, of revealing Compton form factors in the deeply virtual region. We compare this approach to standard least-squares model fitting both for a simplified toy case and for HERMES data.

Deeply virtual Compton scattering; Generalized parton distributions; neural networks

Rad je prezentiran na skupu Ringberg Workshop New Trends in HERA Physics 2011., održanom od 25.–28.09.2011., Ringberg Castle Tegernsee, Njemačka.