Machine Learning Estimators for Lattice QCD Observables
arXiv:1807.05971 · doi:10.1103/PhysRevD.100.014504
Abstract
A novel technique using machine learning (ML) to reduce the computational cost of evaluating lattice quantum chromodynamics (QCD) observables is presented. The ML is trained on a subset of background gauge field configurations, called the labeled set, to predict an observable from the values of correlated, but less compute-intensive, observables calculated on the full sample. By using a second subset, also part of the labeled set, we estimate the bias in the result predicted by the trained ML algorithm. A reduction in the computational cost by about is demonstrated for two different lattice QCD calculations using the Boosted decision tree regression ML algorithm: (1) prediction of the nucleon three-point correlation functions that yield isovector charges from the two-point correlation functions, and (2) prediction of the phase acquired by the neutron mass when a small Charge-Parity (CP) violating interaction, the quark chromoelectric dipole moment interaction, is added to QCD, again from the two-point correlation functions calculated without CP violation.
arXiv comment8 pages, 5 figures
References in corpus (23)
- Scikit-learn: Machine Learning in Python
- Searching for Exotic Particles in High-Energy Physics with Deep Learning
- Electric dipole moments as probes of new physics
- The Chroma Software System for Lattice QCD
- Highly Improved Staggered Quarks on the Lattice, with Applications to Charm Physics
- Electric Dipole Moments of Nucleons, Nuclei, and Atoms: The Standard Model and Beyond
- Lattice QCD ensembles with four flavors of highly improved staggered quarks
- Probing Novel Scalar and Tensor Interactions from (Ultra)Cold Neutrons to the LHC
- Effective noise reduction techniques for disconnected loops in Lattice QCD
- A new class of variance reduction techniques using lattice symmetries
- Adaptive multigrid algorithm for the lattice Wilson-Dirac operator
- Axial, Scalar and Tensor Charges of the Nucleon from 2+1+1-flavor Lattice QCD
- Isovector Charges of the Nucleon from 2+1+1-flavor Lattice QCD
- Machine Learning of Explicit Order Parameters: From the Ising Model to SU(2) Lattice Gauge Theory
- Neutron Electric Dipole Moment and Tensor Charges from Lattice QCD
- Low Energy Precision Test of Supersymmetry
- Nucleon Charges and Electromagnetic Form Factors from 2+1+1-Flavor Lattice QCD
- Enhanced Higgs to Searches with Deep Learning
- Dimension-5 CP-odd operators: QCD mixing and renormalization
- Deep Learning Beyond Lefschetz Thimbles
- Machine learning action parameters in lattice quantum chromodynamics
- Controlling Excited-State Contamination in Nucleon Matrix Elements
- Neutron and proton electric dipole moments from domain-wall fermion lattice QCD
Cited by in corpus (17)
- Machine Learning in Nuclear Physics
- Nuclear matrix elements from lattice QCD for electroweak and beyond-Standard-Model processes
- Stochastic normalizing flows as non-equilibrium transformations
- Status and Future Perspectives for Lattice Gauge Theory Calculations to the Exascale and Beyond
- Learning Lattice Quantum Field Theories with Equivariant Continuous Flows
- Machine-learning physics from unphysics: Finding deconfinement temperature in lattice Yang-Mills theories from outside the scaling window
- Machine-Learning Prediction for Quasi-PDF Matrix Elements
- A Neural Networks study of the phase transitions of Potts model
- Machine learning phases and criticalities without using real data for training
- Machine learning spectral functions in lattice QCD
- A machine learning approach for efficient multi-dimensional integration
- Lossy compression of statistical data using quantum annealer
- A regression algorithm for accelerated lattice QCD that exploits sparse inference on the D-Wave quantum annealer
- Equivariance and generalization in neural networks
- Constructing and Machine Learning Calabi-Yau Five-folds
- Predicting the Masses of Exotic Hadrons with Data Augmentation Using Multilayer Perceptron
- Nuclear correlation functions using first-principle calculations of lattice quantum chromodynamics