Bayesian Optimization for machine learning algorithms in the context of Higgs searches at the CMS experiment

TL;DR

This paper applies Bayesian Optimization to tune hyperparameters of machine learning algorithms used for energy regression of photons and electrons in Higgs searches at the CMS experiment, enhancing the accuracy of particle energy estimation.

Contribution

It introduces the use of Bayesian Optimization for hyperparameter tuning in high-energy physics machine learning applications, specifically for energy regression in the CMS experiment.

Findings

Improved energy estimation accuracy for photons and electrons.

Enhanced hyperparameter tuning efficiency using Bayesian Optimization.

Potential for better Higgs search sensitivity.

Abstract

Machine Learning algorithms, such as Boosted Decisions Trees and Deep Neural Network, are widely used in High-Energy-Physics. The aim of this study is to apply Bayesian Optimization to tune the hyperparameters used in a machine learning algorithm. This algorithm performs an energy regression process on photons and electrons detected in the electromagnetic calorimeter at the Compact Muon Solenoid experiment operating at the Large Hadron Collider at CERN. The goal of this algorithm is to estimate the energy of photons and electrons created during the collisions in the Compact Muon Solenoid, from the measured energy.