Cyclotron Radiation Emission Spectroscopy Signal Classification with Machine Learning in Project 8

TL;DR

This paper presents a machine learning-based classification scheme for complex signals in Cyclotron Radiation Emission Spectroscopy, aiming to enhance energy spectrum analysis and improve sensitivity in beta decay experiments.

Contribution

It introduces a novel machine learning approach to analyze and classify CRES signals, leveraging physical traits for better frequency reconstruction and experimental sensitivity.

Findings

Enhanced signal classification accuracy

Improved cyclotron frequency reconstruction

Potential for increased experimental sensitivity

Abstract

The Cyclotron Radiation Emission Spectroscopy (CRES) technique pioneered by Project 8 measures electromagnetic radiation from individual electrons gyrating in a background magnetic field to construct a highly precise energy spectrum for beta decay studies and other applications. The detector, magnetic trap geometry, and electron dynamics give rise to a multitude of complex electron signal structures which carry information about distinguishing physical traits. With machine learning models, we develop a scheme based on these traits to analyze and classify CRES signals. Understanding and proper use of these traits will be instrumental to improve cyclotron frequency reconstruction and help Project 8 achieve world-leading sensitivity on the tritium endpoint measurement in the future.