Deep-learning jets with uncertainties and more

Precision theoretical predictions for high multiplicity scattering rely on the evaluation of increasingly complicated scattering amplitudes which come with an extremely high CPU cost. For state-of-the-art processes this can cause technical bottlenecks in the production of fully differential distributions. In this article we explore the possibility of using neural networks to approximate multi-variable scattering amplitudes and provide efficient inputs for Monte Carlo integration. We focus on QCD corrections to e + e − → jets up to one-loop and up to five jets. We demonstrate reliable interpolation when a series of networks are trained to amplitudes that have been divided into sectors defined by their infrared singularity structure. Complete simulations for one-loop distributions show speed improvements of at least an order of magnitude over a standard approach.

Section: Uncertainty Analysismentioning

confidence: 99%

Using neural networks for efficient evaluation of high multiplicity scattering amplitudes

Badger

Bullock

2020

“…While the benchmark points representing a cluster are isolated points in the parameter space, the procedure we propose here allows us to associate certain shapes more straightforwardly with distinct regions in the parameter space. The application of machine learning techniques in high energy physics, in particular to constrain the EFT/new physics parameter space, has been brought forward already some time ago [68][69][70][71], with successful applications in jet and top quark identification [72][73][74][75][76][77][78][79][80][81], new physics searches [70,71,[82][83][84][85][86][87][88][89][90] and PDFs [91]. Shape analysis with machine learning has been applied already to constrain anomalous Higgs-vector boson couplings in HZ production [92].…”

Section: Introductionmentioning

confidence: 99%

Exploring anomalous couplings in Higgs boson pair production through shape analysis

Capozi

Heinrich

2020

We classify shapes of Higgs boson pair invariant mass distributions m hh , calculated at NLO with full top quark mass dependence, and visualise how distinct classes of shapes relate to the underlying coupling parameter space. Our study is based on a five-dimensional parameter space relevant for Higgs boson pair production in a non-linear Effective Field Theory framework. We use two approaches: an analysis based on predefined shape types and a classification into shape clusters based on unsupervised learning. We find that our method based on unsupervised learning is able to capture shape features very well and therefore allows a more detailed study of the impact of anomalous couplings on the m hh shape compared to more conventional approaches to a shape analysis.

“…The current ML revolution moves away from low-dimensional projections and exploit deep neural network (NN) to perform classification. Early progress was made on supervised classification of known particles [8][9][10][11][12] to the point where powerful network architectures are now available [13][14][15][16][17] and the focus lies on improving the stability [18][19][20][21][22] of these approaches.…”

Section: Jhep09(2020)195mentioning

confidence: 99%

Towards machine learning analytics for jet substructure

Kasieczka¹,

Marzani²,

Soyez³

et al. 2020

Self Cite

The past few years have seen a rapid development of machine-learning algorithms. While surely augmenting performance, these complex tools are often treated as black-boxes and may impair our understanding of the physical processes under study. The aim of this paper is to move a first step into the direction of applying expert-knowledge in particle physics to calculate the optimal decision function and test whether it is achieved by standard training, thus making the aforementioned black-box more transparent. In particular, we consider the binary classification problem of discriminating quark-initiated jets from gluon-initiated ones. We construct a new version of the widely used N-subjettiness, which features a simpler theoretical behaviour than the original one, while maintaining, if not exceeding, the discrimination power. We input these new observables to the simplest possible neural network, i.e. the one made by a single neuron, or perceptron, and we analytically study the network behaviour at leading logarithmic accuracy. We are able to determine under which circumstances the perceptron achieves optimal performance. We also compare our analytic findings to an actual implementation of a perceptron and to a more realistic neural network and find very good agreement.