Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2211.02834

Multi-variable Integration with a Neural Network

Daniel Maître·Roi Santos-Mateos
View original source ↗Read source document ↗Export metadata ↓

Abstract

In this article we present a method for automatic integration of parametric integrals over the unit hypercube using a neural network. The method fits a neural network to the primitive of the integrand using a loss function designed to minimize the difference between multiple derivatives of the network and the function to be integrated. We apply this method to two example integrals resulting from the sector decomposition of a one-loop and two-loop scalar integrals. Our method can achieve per-mil and percent accuracy for these integrals over a range of invariant values. Once the neural network is fitted, the evaluation of the integral is between 40 and 125 times faster than the usual numerical integration method for our examples, and we expect the speed gain to increase with the complexity of the integrand.

Explore related subjects

hep-ph

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Daniel Maître, Roi Santos-Mateos. 2023-03-13. Multi-variable Integration with a Neural Network. https://doi.org/10.1007/jhep03(2023)221

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Stochastic Ultralight Dark Matter Fluctuations in Pulsar Timing Arrays

Metric perturbations induced by ultralight dark matter (ULDM) fields have long been identified as a potential target for pulsar timing array (PTA) observations. Previous works have focused on the coherent oscillation of metric perturbations at the characteristic frequency set by the ULDM mass. In this work, we show that ULDM fields source low-frequency stochastic metric fluctuations and that these low-frequency fluctuations can produce distinctive detectable signals in PTA data. Using the NANOGrav 12.5-year data set and synthetic data sets mimicking present and future PTA capabilities, we show that the current and future PTA observations provide the strongest probe of ULDM density within the solar system for masses in the range of $10^{-18}\;{\rm eV}-10^{-16}\;{\rm eV}$.

hep-ph↗
arXiv2026

On the Origin of QCD Collectivity in High-Multiplicity Jets: A Transport Model Study

The CMS Collaboration has observed an enhancement of elliptic azimuthal anisotropy ($v^{\ast}_2$) in high-multiplicity jets. To investigate its microscopic origin, we employ a hybrid transport model that couples jets generated with \textsc{PYTHIA~8} to partonic and hadronic rescattering. The analysis is performed in the jet frame, where the jet momentum defines the longitudinal axis. We characterize the initial-state geometry using the eccentricity vectors of shower partons and quantify the geometric response by correlating them with the final-state flow vectors of hadrons. By systematically varying the partonic and hadronic interactions, we find that the anisotropy enhancement is dominated by hadronic rescattering in the present model and increases with the initial eccentricity. We further classify jets using the Soft Drop variable $z_gθ_g^β$ and show that this momentum-space substructure variable is sensitive to the initial coordinate-space geometry, although the predicted substructure dependence differs from the current CMS measurement. These results support a geometry--response mechanism for collective behavior inside jets and establish jet substructure as a promising experimental handle on the initial geometry. They also motivate models that treat parton branching and transport interactions concurrently.

hep-ph↗
arXiv2026

Comprehensive effective field theory framework for coherent elastic neutrino-nucleus scattering

Coherent elastic neutrino-nucleus scattering (CE$ν$NS) stands out as a pivotal process for precision tests of the Standard Model electroweak sector, investigations of neutrino properties, and searches for new physics. Recent experimental measurements by COHERENT, CONUS+, and ton-scale xenon detectors--including PandaX-4T and XENONnT--underscore the need for a systematic theoretical framework to bridge high-energy physics scenarios with low-energy observational data. In this work, we develop a comprehensive end-to-end effective field theory (EFT) framework for CE$ν$NS, encompassing the complete energy scale hierarchy spanning the ultraviolet regime down to the nuclear sector. We consider the low-energy EFT (LEFT) operators up to dimension 8, incorporating their QCD renormalization group running effects, and employ the systematic spurion method to achieve matching between these operators and the chiral Lagrangian. A full power counting analysis is performed, extending to nuclear response functions, which evaluates contributions from LEFT operators up to dimension 8 while accounting for the nucleon number enhancement effect intrinsic to CE$ν$NS. Moreover, we match the relevant LEFT operators for CE$ν$NS onto operators up to dimension 8 within the Standard Model EFT. By also providing their complete tree-level ultraviolet completions, this procedure establishes a consistent top-down theoretical workflow. Leveraging a broad suite of CE$ν$NS experimental data, this framework enables a combined analysis to extract constraints on the scales of EFT operators and neutrino non-standard interaction parameters.

hep-ph↗