README.md

In hsansford1/higgsboson:

higgsboson

Package for SM1/SC1 group project. Used to analyse the simulated data from the ATLAS experiment at CERN to optimise the analysis of the Higgs Boson.

Installation

This package can be installed using devtools and the command

devtools::install_github("hsansford1/higgsboson")

Introduction

The Atlas experiment at CERN provided simulated data that was used to optimise the analysis of the Higgs boson. In the Large Hadron Collider (LHC), proton bunches are accelerated in both directions on a circular trajectory. This results in some of the protons colliding as the bunches cross the ATLAS detector (called an event), which produces hundreds of millions of proton-proton collisions per second. The particles resulting from each event are detected by sensors and, from this raw data, certain real-valued features are estimated.

Most of the uninteresting events (called the background) are discarded using a real-time multi-stage cascade classifier. However, many of the remaining events represent known processes (also called background). Our aim is to find the region of the feature space in which there is a significant excess of events (called signal) compared to what known background processes can explain.

Data

To load the data, use the function data with either training or test as input, i.e.

data(training)
data(test)

Both datasets are very large with 30 feature columns, a weight column and a label column. The training data has 250,000 rows and the test data has 550,000 rows.

hsansford1/higgsboson documentation built on Jan. 22, 2022, 4:34 a.m.