Identifying hot subdwarf stars from photometric data using Gaussian mixture model and graph neural network

TL;DR

This paper introduces a novel machine learning approach combining Gaussian mixture models and graph neural networks to identify hot subdwarf stars from photometric data, significantly improving detection accuracy.

Contribution

The paper presents a new method integrating Gaussian mixture models and graph neural networks for hot subdwarf star identification in large photometric datasets.

Findings

Achieved high recall, precision, and F1 scores on multiple datasets.

Selected approximately 6,000 candidate stars most similar to hot subdwarfs.

Demonstrated effectiveness of the combined GMM and GNN approach in stellar classification.

Abstract

Hot subdwarf stars are very important for understanding stellar evolution, stellar astrophysics, and binary star systems. Identifying more such stars can help us better understand their statistical distribution, properties, and evolution. In this paper, we present a new method to search for hot subdwarf stars in photometric data (b, y, g, r, i, z) using a machine learning algorithm, graph neural network, and Gaussian mixture model. We use a Gaussian mixture model and Markov distance to build the graph structure, and on the graph structure, we use a graph neural network to identify hot subdwarf stars from 86 084 stars, when the recall, precision, and f1 score are maximized on the original, weight and synthetic minority oversampling technique datasets. Finally, from 21 885 candidates, we selected approximately 6 000 stars that were the most similar to the hot subdwarf star.