Influence of Fake Track Rejection on Large-Radius Unified Flow Object Jet Reconstruction and the Development of a Deep Neural Network to Optimize High-p$_T$ Cluster Energy Extrapolation

This report summarizes my work studying the influence of fake tracks on jet reconstruction using Unified Flow Objects as inputs to the anti-k$_T$ algorithm for the ATLAS Collaboration. High-p$_T$ and dense environments can lead to degraded performance in reconstruction, resulting in an increase in fake tracks, which are incorrectly reconstructed tracks. We studied the impact of fake tracks on pile-up stability and jet mass scale and resolution for boosted W jets. The document also summarizes my progress in developing a deep neural network to predict the energy a charged particle will deposit in a cluster. The study aims to begin to understand the complicated nature of the relationship between matched tracks and clusters at high-p$_T$ and how a neural network may be used to model the energy in the clusters matched to charged particles.