Jet substructure classification in high-energy physics with deep neural networks

Abstract

At the extreme energies of the Large Hadron Collider, massive particles can be produced at such high velocities that their hadronic decays are collimated and the resulting jets overlap. Deducing whether the substructure of an observed jet is due to a low-mass single particle or due to multiple decay objects of a massive particle is an important problem in the analysis of collider data. Traditional approaches have relied on expert features designed to detect energy deposition patterns in the calorimeter, but the complexity of the data make this task an excellent candidate for the application of machine learning tools. The data collected by the detector can be treated as a twodimensional image, lending itself to the natural application of image classification techniques. In this work, we apply deep neural networks with a mixture of locally-connected and fully-connected nodes. Our experiments demonstrate that without the aid of expert features, such networks match or modestly outperform the current state-of-the-art approach for discriminating between jets from single hadronic particles and overlapping jets from pairs of collimated hadronic particles, and that such performance gains persist in the presence of pileup interactions.

Jet flavor classification in high-energy physics with deep neural networks

“…Recent applications of deep learning to similar problems in high-energy physics [6][7][8][9], combined with the lack of a clear analytical theory to provide dimensional reduction without loss of information, suggests that deep learning techniques applied to the lower-level higherdimensional data could yield improvements in the performance of jet-flavor classification algorithms. General methods for designing and applying recurrent and recursive neural networks to problems with data of variable size or structure have been developed in Refs.…”

What is the machine learning?

“…In what follows, as we study planing we will also utilize a technique (see [2][3][4][5]11,12]) which we refer to as "saturation," that compares a network trained on only low-level inputs with networks trained after adding higher-level variables. Saturation provides a tool to ensure that our networks are sufficiently deep, by checking that the new network's performance does not improve by much.…”

“…The authors of [2][3][4][5] emphasized the ability of deep learning to outperform physics inspired high-level variables. We use the "uniform phase space" scheme to flatten discriminating variables, which was introduced in [6] to quantify the information learned by deep neural networks.…”

Deep-learning top taggers or the end of QCD?

“…An example of such an approach are wavelets, describing patterns of hadronic weak boson decays [20,21]. Even more generally, we can apply image recognition techniques to the two-dimensional azimuthal angle vs rapidity plane, for example searching for hadronic decays of weak bosons [22][23][24][25] or top quarks [26]. The same techniques can be applied to separate quark-like and gluon-like jets [27].…”