The use of Machine Learning techniques for the identification of pi0s in the MPC-EX Analysis of d+Au collisions at sqrt(s_NN) = 200 GeV

The Muon Piston Calorimeter Extension (MPC-EX), is a combination of high resolution charged particle tracker and electromagnetic pre-shower detector, which sits in front of a PbW crystal calorimeter located at forward rapidities ($3.1 < |\eta|< 3.8$) in the PHENIX detector. It is uniquely positioned in d+Au collisions to measure phenomena related to low-x partons in the target nucleus and high-x partons in the projectile hadron [1]. The eight Si-W layers have a spatial resolution of 1.8 mm and work to capture the structure of shower profile. In recent years, machine learning techniques have achieved great success in many fields, particularly in image recognition. The shower structures taken by the MPC-EX can be viewed as images. In this talk, we apply these techniques to identify neutral $\pi^{0}$s in the MPC-EX at high energies where the showers overlap, and distinguish them from the pileup of electromagnetic showers from other sources.