Targeting multi-loop integrals with neural networks
Ramon Winterhalder, Vitaly Magerya, Emilio Villa, Stephen Jones, Matthias Kerner, Anja Butter, Gudrun Heinrich, Tilman Plehn
Abstract
Numerical evaluations of Feynman integrals often proceed via a deformation of the integration contour into the complex plane. While valid contours are easy to construct, the numerical precision for a multi-loop integral can depend critically on the chosen contour. We present methods to optimize this contour using a combination of optimized, global complex shifts and a normalizing flow. They can lead to a significant gain in precision.
Topics & Concepts
Methods of contour integrationLoop (graph theory)Numerical integrationArtificial neural networkComputer scienceConstruct (python library)AlgorithmComplex planePlane (geometry)Numerical analysisContour lineFlow (mathematics)MathematicsApplied mathematicsArtificial intelligenceMathematical analysisGeometryPhysicsCombinatoricsMeteorologyProgramming languageNumerical Methods and AlgorithmsModel Reduction and Neural NetworksComputational Physics and Python Applications