Please visit Jefferson Lab Event Policies and Guidance before planning your next event: https://www.jlab.org/conference_planning.

Sep 24 – 29, 2023
US/Eastern timezone

Deep Neural Network-Based Reconstruction to extract unpolarized Drell-Yan Asymmetries and the Boer-Mulders Function from SeaQuest E906 Data

Sep 27, 2023, 10:15 AM
15m
Meeting Room 1-2 (Durham Convention Center)

Meeting Room 1-2

Durham Convention Center

Talk 3D Structure of the Nucleon: TMDs 3D Structure of the Nucleon: TMDs

Speaker

Arthur Conover (University of Virginia)

Description

The Boer-Mulders function is a transverse momentum distribution that describes the net polarization of partons within an unpolarized nucleon. A non-zero Boer-Mulders function suggests a handedness of the nucleon and gives rise to a measurable azimuthal asymmetry in Drell-Yan scattering. We suggest a novel approach utilizing DNN-based reconstruction techniques to extract unpolarized Drell-Yan asymmetries from SeaQuest E906 data, which can be used to extract the Boer-Mulders function. This reconstruction technique could be significantly faster than existing methods, potentially allowing us to lower statistical error on the angular-dependence coefficients $\lambda$, $\mu$, and $\nu$. With lower statistical error, we can calculate the value of these constants in more bins, allowing for more general extraction of the Boer-Mulders function, which is highly sensitive to the $p_T$ dependence of $\nu$. SeaQuest at Fermilab was a fixed-target experiment designed to detect the Drell-Yan process in $p+p$ and $p+d$ reactions. We discuss the steps involved in our DNN reconstruction process, including data preprocessing, network architecture design, and training strategies, while addressing challenges such as background estimation, and systematic uncertainties.

Primary author

Arthur Conover (University of Virginia)

Presentation materials