Optimization of convolutional neural networks for background suppression in the PandaX-III experiment

TL;DR

This paper enhances CNN-based background suppression in the PandaX-III experiment by optimizing hyperparameters, notably channel number, leading to a 70% improvement in signal-background discrimination for neutrinoless double beta decay detection.

Contribution

It introduces an optimized CNN model with hyperparameter tuning, significantly improving background discrimination in the PandaX-III experiment.

Findings

70% improvement in discrimination significance

Optimized hyperparameters enhance CNN performance

Effective background suppression for neutrinoless double beta decay

Abstract

The tracks recorded by a gaseous detector provide a possibility for charged particle identification. For searching the neutrinoless double beta decay events of 136Xe in the PandaX-III experiment, we optimized the convolutional neural network based on the Monte Carlo simulation data to improve the signal-background discrimination power. EfficientNet is chosen as the baseline model and the optimization is performed by tuning the hyperparameters. In particular, the maximum discrimination power is achieved by optimizing the channel number of the top convolutional layer. In comparison with our previous work, the significance of discrimination has been improved by about 70%.