SciPost logo

SciPost Submission Page

Comparing Machine Learning and Interpolation Methods for Loop-Level Calculations

by Ibrahim Chahrour, James D. Wells

This Submission thread is now published as

Submission summary

Authors (as registered SciPost users): Ibrahim Chahrour
Submission information
Preprint Link: https://arxiv.org/abs/2111.14788v3  (pdf)
Date accepted: 2022-05-10
Date submitted: 2022-03-30 03:11
Submitted by: Chahrour, Ibrahim
Submitted to: SciPost Physics
Ontological classification
Academic field: Physics
Specialties:
  • High-Energy Physics - Phenomenology
Approach: Computational

Abstract

The need to approximate functions is ubiquitous in science, either due to empirical constraints or high computational cost of accessing the function. In high-energy physics, the precise computation of the scattering cross-section of a process requires the evaluation of computationally intensive integrals. A wide variety of methods in machine learning have been used to tackle this problem, but often the motivation of using one method over another is lacking. Comparing these methods is typically highly dependent on the problem at hand, so we specify to the case where we can evaluate the function a large number of times, after which quick and accurate evaluation can take place. We consider four interpolation and three machine learning techniques and compare their performance on three toy functions, the four-point scalar Passarino-Veltman $D_0$ function, and the two-loop self-energy master integral $M$. We find that in low dimensions ($d = 3$), traditional interpolation techniques like the Radial Basis Function perform very well, but in higher dimensions ($d=5, 6, 9$) we find that multi-layer perceptrons (a.k.a neural networks) do not suffer as much from the curse of dimensionality and provide the fastest and most accurate predictions.

List of changes

- Title changed from "Function Approximation for High-Energy Physics: Comparing Machine Learning and Interpolation Methods" to "Comparing Machine Learning and Interpolation Methods for Loop-Level Calculations"
- Added references [7], [8], [16], [17], [18]
- Clarified the results of previous papers in the introduction, in particular the gain in speed of references [9] and [10]
- Clarified the current state of Machine Learning methods in the field of high-energy physics in the introduction

Published as SciPost Phys. 12, 187 (2022)


Reports on this Submission

Report #2 by Anonymous (Referee 2) on 2022-4-19 (Invited Report)

Report

I am happy to recommend publication of the revised version of this paper.

  • validity: -
  • significance: -
  • originality: -
  • clarity: -
  • formatting: -
  • grammar: -

Report #1 by Simon Badger (Referee 1) on 2022-4-5 (Invited Report)

Report

I was happy to read the responses to both referee reports and am happy with the improvements that have been made to the article.

I can recommend it for publication in its current form.

  • validity: -
  • significance: -
  • originality: -
  • clarity: -
  • formatting: -
  • grammar: -

Login to report or comment