Classification of Signal Versus Background in High-Energy Physics Using Deep Neural Networks

Abstract

High-energy physics is a fertile area for applied research in machine learning and deep learning. The Large hadron collider generates humongous amount of data by colliding hadrons at very high velocities and recording the events by various detectors. The data about the events are extensively used by machine learning algorithms to classify particles and also find new exotic particles. Deep learning is a specialization of artificial intelligence and machine learning that uses multi-layered artificial neural networks to excel in activities such as detecting the object, recognizing the speech, etc. Classical Techniques such as shallow neural networks have limitations to study the complex non-linear functions of the inputs. Deep learning programs should have access to large amounts of training data and processing power to attain an acceptable level of accuracy. These techniques have made significant progress in the classification metric which uses the best new approaches without the manual assistance.