Towards Understanding the Milky Way's Matter Field and Dynamical Accretion History based on AI-GS3 Hunter

TL;DR

GS3 Hunter is a deep learning tool that identifies stellar substructures and streams in the Milky Way, revealing complex accretion history and halo assembly through analysis of large astronomical datasets.

Contribution

It introduces a novel combination of Siamese Neural Networks and K-means clustering for stellar substructure detection in galactic data.

Findings

Recovered known stellar groups like Thamnos and GSE

Discovered GSE comprises four distinct components, indicating multiple accretion events

Uncovered new substructures in stellar kinematic data

Abstract

We present GS3 Hunter (Galactic-Seismology Substructures and Streams Hunter), a novel deep-learning method that combines Siamese Neural Networks and K-means clustering to identify substructures and streams in stellar kinematic data. Applied to Gaia EDR3 and GALAH DR3, it recovers known groups (e.g., Thamnos, Helmi, GSE, Sequoia) and, with DESI dataset, reveals that GSE consists of four distinct components (GSH-GSH1 through GSE-GSH4), implying a multi-event accretion origin. Tests on LAMOST K-giants recover Sagittarius, Hercules-Aquila, and Virgo Overdensity, while also uncovering new substructures. Validation with FIRE simulations shows good agreement with previous results. GS3 Hunter thus offers a powerful tool to understand the Milky Way's halo assembly and tidal history.