Age Determination of LAMOST Red Giant Branch stars based on the Gradient Boosting Decision Tree method

TL;DR

This paper develops a machine learning approach using Gradient Boosting Decision Trees to estimate ages of Red Giant Branch stars from LAMOST data, achieving about 15-30% uncertainty and demonstrating its applicability to large stellar samples.

Contribution

The study introduces a GBDT-based method for stellar age estimation that leverages spectroscopic parameters, providing a new tool for large-scale stellar population analysis.

Findings

Median relative error of 11.6% in age estimation.

Systematic differences compared to other methods.

Age uncertainties range from 15% to 30%.

Abstract

In this study we estimate the stellar ages of LAMOST DR8 Red Giant Branch (RGB) stars based on the Gradient Boosting Decision Tree algorithm (GBDT). We used 2,643 RGB stars extracted from the APOKASC-2 astero-seismological catalog as training data-set. After selecting the parameterses ([$\alpha$/Fe], [C/Fe], T$_{eff}$, [N/Fe], [C/H], log g) highly correlated with age using GBDT, we apply the same GBDT method to the new catalog of more than 590,000 stars classified as RGB stars. The test data-set shows that the median relative error is around 11.6$\%$ for the method. We also compare the predicted ages of RGB stars with other studies (e.g., based on APOGEE), and find systematic differences. The final uncertainty is about 15 to 30$\%$ compared to open clusters' ages. Then we present the spatial distribution of the RGB sample having an age determination, which could recreate the expected result, and discuss systematic biases. All these diagnostics show that one can apply the GBDT method to other stellar samples to estimate atmospheric parameters and age.