# From the Bottom to the Top -- Reconstruction of $t\bar{t}$ Events with Deep Learning

## TL;DR

This paper presents a deep learning approach for reconstructing top-quark pair-production events, significantly improving jet-parton assignment accuracy compared to traditional methods, with a workflow for hyperparameter optimization adaptable to experimental setups.

## Contribution

Introduces a deep neural network for $t\bar{t}$ event reconstruction, outperforming kinematic fits and providing a reproducible hyperparameter optimization workflow.

## Key findings

- Significant improvement in jet-to-parton assignment accuracy.
- Enhanced reconstruction of top-quark kinematic properties.
- Workflow enables adaptation to various detector simulations.

## Abstract

The reconstruction of top-quark pair-production ($t\bar{t}$) events is a prerequisite for many top-quark measurements. We use a deep neural network, trained with Monte-Carlo simulated events, to reconstruct $t\bar{t}$ decays in the lepton+jets final state. Comparing our approach to a widely-used kinematic fit, we find significant improvements in the correct assignment of jets to the partons from the decay, and we study the reconstruction performance of several kinematic top-quark properties. We document our workflow for the optimisation of the hyperparameters of the deep neural network. This workflow can be followed by experimental collaborations to retrain the network taking into account their detailed detector simulations.