The European Space Agency's Gaia mission released its third data set, providing improved measurements for 1.8 billion stars, offering a vast amount of data for astronomers. However, handling this data presents significant challenges. The researchers addressed this by utilizing machine learning to estimate key stellar properties from Gaia's spectrophotometric data. The model, trained on high-quality data from 8 million stars, achieved highly reliable predictions with minimal uncertainty.

"The underlying technique, called extreme gradient-boosted trees allows to estimate precise stellar properties, such as temperature, chemical composition, and interstellar dust obscuration, with unprecedented efficiency. The developed machine learning model, SHBoost, completes its tasks, including model training and prediction, within four hours on a single GPU - a process that previously required two weeks and 3000 high-performance processors," said Arman Khalatyan from AIP, lead author of the study. "The machine-learning method is thus significantly reducing computational time, energy consumption, and CO2 emission." This marks the first successful application of this technique to stars of all types simultaneously.

The model leverages high-quality spectroscopic data from smaller stellar surveys and applies it to Gaia's extensive third data release (DR3), estimating key stellar parameters using photometric and astrometric data alongside Gaia's low-resolution XP spectra. "The high quality of the results reduces the need for additional resource-intensive spectroscopic observations when looking for good candidates to be picked-up for further studies, such as rare metal-poor or super-metal rich stars, crucial for understanding the earliest phases of the Milky Way formation," added Cristina Chiappini from AIP. This approach is also pivotal for preparing future multi-object spectroscopy observations, such as the 4MOST project's 4MIDABLE-LR survey at the European Southern Observatory in Chile.

"The new model approach provides extensive maps of the Milky Way's overall chemical composition, corroborating the distribution of young and old stars. The data shows the concentration of metal-rich stars in the Galaxy's inner regions, including the bar and bulge, with an enormous statistical power," said Friedrich Anders from ICCUB.

The team used the model to chart young, massive stars across the galaxy, highlighting distant, understudied star-forming regions. The data also revealed "stellar voids," areas with few young stars, and pointed out where the three-dimensional distribution of interstellar dust is still not well understood.

As Gaia continues to gather data, machine-learning models like SHBoost are becoming crucial tools for quickly and sustainably processing large datasets. This success highlights machine learning's potential to revolutionize data analysis in astronomy and other fields, promoting more sustainable research practices.

Research Report:Transferring spectroscopic stellar labels to 217 million Gaia DR3 XP stars with SHBoost

Related Links
Institute of Cosmos Sciences of the University of Barcelona
Stellar Chemistry, The Universe And All Within It