Predicting Feynman periods in φ4-theory

Shaban, Kimia
DOI: https://doi.org/10.1007/jhep11(2024)038
IF: 6.379
2024-11-08
Journal of High Energy Physics
Abstract:We present efficient data-driven approaches to predict the value of subdivergence-free Feynman integrals (Feynman periods) in φ 4 -theory from properties of the underlying Feynman graphs, based on a statistical examination of almost 2 million graphs. We find that the numbers of cuts and cycles determines the period to better than 2% relative accuracy. Hepp bound and Martin invariant allow for even more accurate predictions. In most cases, the period is a multi-linear function of the properties in question. Furthermore, we investigate the usefulness of machine-learning algorithms to predict the period. When sufficiently many properties of the graph are used, the period can be predicted with better than 0.05% relative accuracy.
physics, particles & fields
What problem does this paper attempt to address?