Robust Neural Particle Identification Models

TL;DR

This paper introduces a robust neural network approach for particle identification at the LHCb detector that mitigates biases from training data, improving efficiency across various decay modes.

Contribution

It proposes a novel Common Specific Decomposition method to disentangle decay-specific features, enhancing neural PID robustness against training data biases.

Findings

Reduces efficiency degradation in particle identification.

Improves robustness of neural models to training data biases.

Enhances detection performance for diverse decay modes.

Abstract

The volume of data processed by the Large Hadron Collider experiments demands sophisticated selection rules typically based on machine learning algorithms. One of the shortcomings of these approaches is their profound sensitivity to the biases in training samples. In the case of particle identification (PID), this might lead to degradation of the efficiency for some decays not present in the training dataset due to differences in input kinematic distributions. In this talk, we propose a method based on the Common Specific Decomposition that takes into account individual decays and possible misshapes in the training data by disentangling common and decay specific components of the input feature set. We show that the proposed approach reduces the rate of efficiency degradation for the PID algorithms for the decays reconstructed in the LHCb detector.