A Neural Resampler for Monte Carlo Reweighting with Preserved Uncertainties

TL;DR

This paper introduces neural resampling for Monte Carlo event reweighting in collider physics, which preserves statistical uncertainties and scales efficiently to complex phase spaces, improving upon traditional methods.

Contribution

It presents a neural network-based unbinned resampling method that maintains event sample uncertainties, applicable to high-dimensional phase spaces and compatible with existing binned resampling techniques.

Findings

Successfully applied to LHC top quark pair production data.

Preserves both mean observables and their Monte Carlo uncertainties.

Scales well to high-dimensional phase space.

Abstract

Monte Carlo event generators are an essential tool for data analysis in collider physics. To include subleading quantum corrections, these generators often need to produce negative weight events, which leads to statistical dilution of the datasets and downstream computational costs for detector simulation. Building on the recent proposal of a positive resampler method to rebalance weights within histogram bins, we introduce neural resampling: an unbinned approach to Monte Carlo reweighting based on neural networks that scales well to high-dimensional and variable-dimensional phase space. We pay particular attention to preserving the statistical properties of the event sample, such that neural resampling not only maintains the mean value of any observable but also its Monte Carlo uncertainty. This uncertainty preservation scheme is general and can also be applied to binned (non-neural network) resampling. To illustrate our neural resampling approach, we present a case study from the Large Hadron Collider of top quark pair production at next-to-leading order matched to a parton shower.