10/10/2024, 04:05 PM UTC
借助 Gaia 和机器学习更清晰地观察银河系A Sharper Look at the Milky Way with Gaia and Machine Learning
➀ 来自波茨坦莱布尼茨天体物理研究所(AIP)和巴塞罗那宇宙科学研究所(ICCUB)的研究人员使用一种新颖的机器学习模型,高效地处理了盖亚任务中 2170 万颗恒星的观测数据。这些结果与传统确定恒星参数的方法相当。新的方法为在整个银河系中绘制恒星消光和金属丰度等特性提供了激动人心的可能性,有助于理解恒星种群和银河系的构成。➁ 欧洲航天局(ESA)盖亚卫星的第三次数据发布提供了对 18 亿颗恒星改进测量的访问权限,这是研究银河系的大量数据。然而,对如此大的数据集进行高效分析却是一个挑战。现在发表的这项研究调查了基于盖亚的光谱光度数据使用机器学习确定重要恒星特性的方法。该模型在 800 万颗恒星的高质量数据上进行了训练,并实现了具有低不确定性的可靠预测。➂ '极端梯度提升树'这种机器学习技术,以前所未有的效率确定精确的恒星特性,如温度、化学组成和星际尘埃消光。开发的机器学习模型 SHBoost 在单个图形处理器上完成其任务,包括模型训练和预测,仅需四小时 - 而之前这一过程需要两周时间和 3000 台高性能处理器。➀ Researchers from the Leibniz Institute for Astrophysics Potsdam (AIP) and the Institute of Cosmos Sciences at the University of Barcelona (ICCUB) have used a novel machine learning model to process observation data from 217 million stars of the Gaia mission efficiently. The results are comparable to conventional methods for determining stellar parameters. The new approach opens up exciting possibilities for mapping properties like interstellar extinction and metallicity across the Milky Way, contributing to understanding the stellar populations and the structure of our galaxy. ➁ The third data release of the Gaia satellite by the European Space Agency ESA provided access to improved measurements for 1.8 billion stars, a vast amount of data for studying the Milky Way. Efficient analysis of such a large dataset, however, presents a challenge. The study published now investigates the use of machine learning to determine important stellar properties based on Gaia's spectrophotometric data. The model was trained on high-quality data from 8 million stars and achieved reliable predictions with low uncertainties. ➂ The machine learning technique, 'Extreme Gradient-Boosted Trees,' enables the determination of precise stellar properties like temperature, chemical composition, and interstellar dust extinction with unprecedented efficiency. The developed machine learning model, SHBoost, completes its tasks, including model training and prediction, within four hours on a single graphics processor - a process that previously required two weeks and 3000 high-performance processors.
---
本文由大语言模型(LLM)生成,旨在为读者提供半导体新闻内容的知识扩展(Beta)。