• Nuclear, Particle, Astroparticle and Cosmology (NUPAC) Seminars

May 5, 2026 2:00 PM
PAIS 3205

Host:

Michael Gold

Presenter:

Zafer Acar (UNM)

Neutrinos are electrically neutral, extremely light particles whose observed flavor oscillations show they have mass, pointing to physics beyond the Standard Model. One of the most sensitive probes of this new physics is neutrinoless double-beta decay (0νββ), a process in which two electrons are emitted without neutrinos. Observation of this decay would violate lepton number and establish neutrinos as Majorana particles, with important implications for the origin of mass and matter–antimatter asymmetry. This talk presents the motivation and experimental search for 0νββ in the LEGEND experiment, which uses enriched high-purity germanium detectors in a low-background liquid argon environment . A key challenge is reducing and understanding background near the region of interest, addressed through Geant4-based simulations using the REMAGE framework to model detector response and physics processes . In addition, machine learning methods are applied to waveform data using self-supervised contrastive learning to group similar events and separate different ones, helping distinguish signal from background. Neural network models are also used for background classification, and synthetic data techniques such as SMOTE are applied to handle class imbalance and improve training. PCA is used to visualize learned features and better understand event structure. These combined simulation and data-driven approaches improve event classification and background rejection, supporting future sensitivity goals for experiments such as LEGEND-1000.