Machine learning for event selection in high energy physics

  title={Machine learning for event selection in high energy physics},
  author={Shimon Whiteson and Daniel Whiteson},
  journal={Eng. Appl. Artif. Intell.},
The field of high energy physics aims to discover the underlying structure of matter by searching for and studying exotic particles, such as the top quark and Higgs boson, produced in collisions at modern accelerators. Since such accelerators are extraordinarily expensive, extracting maximal information from the resulting data is essential. However, most accelerator events do not produce particles of interest, so making effective measurements requires event selection, in which events producing… Expand
Optimal event selection and categorization in high energy physics, Part 1: Signal discovery
We provide a prescription to train optimal machine-learning-based event selectors and categorizers that maximize the statistical significance of a potential signal excess in high energy physics (HEP)Expand
A novel multi-variate analysis method for searching particles in high energy physics
Based on the construction and approximation of the extreme selection region presented in this paper, the DEFE model is able to be trained efficiently, and extract discriminative features from multiple angles and dimensions, hence the improvement of the selection region of searching new particles in HEP can be achieved. Expand
ThickBrick: optimal event selection and categorization in high energy physics. Part I. Signal discovery
We provide a prescription called ThickBrick to train optimal machine-learning-based event selectors and categorizers that maximize the statistical significance of a potential signal excess in highExpand
Deep Extreme Feature Extraction: New MVA Method for Searching Particles in High Energy Physics
Deep Extreme Feature Extraction can be viewed as a deep ensemble learning scheme that trains a strongly diverse set of neural feature learners without explicitly encouraging diversity and penalizing correlations, and is able to train an ensemble of discriminative feature learners that boosts the overperformance of final prediction. Expand
Generative models for fast simulation
  • S. Vallecorsa
  • Computer Science
  • Journal of Physics: Conference Series
  • 2018
The first three-dimensional images of energy showers in a high granularity calorimeter, obtained using Generative Adversarial Networks, are presented. Expand
FPGA accelerated computing for particle identification in high-energy physics experiments
The key challenges in the recent nuclear and High-Energy Physics (HEP) experiments are high reaction rate and data rate. The data rate, the product of the event size and the reaction rate, rangesExpand
Opportunities in Machine Learning for Particle Accelerators
Machine learning” (ML) is a subfield of artificial intelligence. The term applies broadly to a collection of computational algorithms and techniques that train systems from raw data rather than aExpand
Application of genetic programming for proton-proton interactions
AbstractThe aim of the present work is to use one of the machine learning techniques named the genetic programming (GP) to model the p-p interactions through discovering functions. In our study, GPExpand
Studying total proton-proton cross section collision at large hadron collider using gene expression programming
New technique is presented for modeling total cross section of proton-proton (p-p) collision from low to ultra-high energy regions using gene expression programming (GEP). GEP, as a machine learningExpand
Neural Networks for Modeling and Control of Particle Accelerators
The purpose of this paper is to re-introduce neural networks to the particle accelerator community and report on some work in neural network control that is being conducted as part of a dedicated collaboration between Fermilab and Colorado State University. Expand


Stochastic Optimization for Collision Selection in High Energy Physics
A new approach that uses stochastic optimization techniques to directly search for selectors that minimize statistical uncertainty in the top quark mass measurement is presented, confirming that stochastically optimized selectors have much smaller uncertainty. Expand
Support vector regression as a signal discriminator in high energy physics
Examples of the application of support vector regression to the search for a theoretically predicted particle and to the classification of heavy quark jets are presented, and comparisons to the performance of an artificial neural network are provided. Expand
Measurement of the cross section for tt̄ production in pp̄ collisions using the kinematics of lepton+jets events
We present a measurement of the top pair production cross section in p (p) over bar collisions at root s=1.96 TeV. We collect a data sample with an integrated luminosity of 194 +/- 11 pb(-1) with theExpand
PhysicsGP: A Genetic Programming approach to event selection
A novel multivariate classification technique based on Genetic Programming that optimizes a set of human-readable classifiers with respect to some user-defined performance measure is presented and offers several advantages compared to Neural Networks and Support Vector Machines. Expand
Review of Particle Physics
This biennial Review summarizes much of particle physics. Using data from previous editions, plus 2158 new measurements from 551 papers, we list, evaluate, and average measured properties of gaugeExpand
Top-quark mass measurement from dilepton events at CDF II.
This first application of a matrix-element technique to tt --> bl+ nu(l)bl'- nu'+ nu'(l') decays gives the most precise single measurement of M(t) in dilepton events. Expand
Search for single top quark production at DØ using neural networks
We present a search for electroweak production of single top quarks in approximate to 90 pb(-1) of data collected with the D(null set) detector at the Fermilab Tevatron collider. Using arrays ofExpand
Observation of top quark production in $\bar{p}p$ collisions
We establish the existence of the top quark using a 67 pb^-1 data sample of Pbar-P collisions at Sqrt(s) = 1.8 TeV collected with the Collider Detector at Fermilab (CDF). Employing techniques similarExpand
High-energy physics event generation with PYTHIA 6.1
Pythia version 6 represents a merger of the Pythia 5, Jetset 7 and SPythia programs, with many improvements. It can be used to generate high-energy-physics ‘events’, i.e. sets of outgoing particlesExpand
Measurement of the top-quark mass with dilepton events selected using neuroevolution at CDF.
A measurement of the top-quark mass M_{t} in the dilepton decay channel tt[over] --> bl;{'+} nu_{l};{'}b[over ]l;{-}nu[ over ]_{l}. Expand