Classification of Red Clump Stars and Red Giant Branch Stars Based on Deep Learning Method
To obtain a purer sample of red giant branch stars,the study classify the evolutionary types of o-ver 20 000 stars from APOGEE-2 with deep learning method.The model is trained using 4216 stars from APOKASC-2 as the training set and applied to classify red giant branch and red clump stars.Using 1807 stars from APOKASC-2 as the test set,the result obtained shows that the mean square error,root mean square error,and mean absolute error of the method applied are 4％,20％,and 9％.Applying the model to more than 20 000 stars from APOGEE-2,the mean square error,root mean square error,and mean ab-solute error of the results are 5％,24％,and 8％.Additionally,the analysis is tested with data from LAM-OST DR7,demonstrating a current accuracy of 98％.The accuracy of the method in this study is basically equal to that of some other classification methods after comparison.Therefore,our method can be used in larger sky surveys for the classification of red clump and red giant branch stars.
red giant branchmachine learningstellar parametersasteroseismologyred clump stars
万方数据
