Improvement of $q^2$ resolution in semileptonic decays based on machine learning

TL;DR

This paper introduces a machine learning approach to improve the resolution of the reconstructed invariant mass squared ($q^2$) in semileptonic decays, achieving about 40% better accuracy than traditional methods.

Contribution

It proposes a novel ML-based method using flight vector features and MLP regressors to resolve the quadratic ambiguity in $q^2$ reconstruction.

Findings

ML method improves $q^2$ resolution by ~40%.

Using flight vector features yields best performance.

Applicable to various semileptonic decays.

Abstract

The neutrino closure method is often used to obtain kinematics of semileptonic decays with one unreconstructed particle. The kinematics of decays can be deducted by a two-fold ambiguity with a quadratic equation. To resolve the two-fold ambiguity, a new method based on Machine Learning (ML) is proposed. We study the effect of different sets of features and regressors on the improvement of reconstructed invariant mass squared of $\ell \nu$ system~($q^2$). The result shows that the best performance is obtained by using the flight vector as the features, and the multilayer perceptron (MLP) model as the regressor. Compared with the random choice, the MLP model improves the resolution of reconstructed $q^2$ by $\sim$40\%. Furthermore, the possibility of using this method on various semileptonic decays is shown.