Nuclear Physics & RIKEN Theory Seminar

"Studying 3D structure of proton with neural networks"

Presented by Kresimir Kumericki, University of Zagreb

Friday, February 3, 2012, 2:00 pm — Small Seminar Room, Bldg. 510

3-dimensional quark-gluon structure of fast proton, encoded in generalized parton distributions (GPDs), is both a testing ground for QCD and an important input into analysis of proton-proton collisions, such as those at LHC. However, extraction of GPDs from experimental data is fraught with uncertainties. After describing the neural network method of data analysis, it will be shown how this method enables elegant and reliable estimation of relevant structure functions. This will be applied to extraction of GPD H from HERMES data on Deeply Virtual Compton Scattering (DVCS).

Hosted by: Bjoern Schenke

7909  |  INT/EXT  |  Events Calendar