Rejecting noise in Baikal-GVD data with neural networks

TL;DR

This paper presents a neural network approach, based on a U-net architecture, for effectively distinguishing signal hits from background noise in Baikal-GVD underwater neutrino data, achieving high purity and efficiency.

Contribution

It introduces a novel neural network method utilizing temporal structure for noise rejection in Baikal-GVD data, outperforming traditional algorithmic approaches.

Findings

Achieves 99% signal purity and 96% survival efficiency on simulated data.

Demonstrates the effectiveness of neural networks in underwater neutrino detection.

Discusses potential for alternative neural network architectures.

Abstract

Baikal-GVD is a large ($\sim$1 km$^3$) underwater neutrino telescope installed in the fresh waters of Lake Baikal. The deep lake water environment is pervaded by background light, which is detectable by Baikal-GVD's photosensors. We introduce a neural network for an efficient separation of these noise hits from the signal ones, stemming from the propagation of relativistic particles through the detector. The model has a U-net-like architecture and employs temporal (causal) structure of events. The neural network's metrics reach up to 99\% signal purity (precision) and 96\% survival efficiency (recall) on Monte-Carlo simulated dataset. We compare the developed method with the algorithmic approach to rejecting the noise and discuss other possible architectures of neural networks, including graph-based ones.