• <tr id='rIDUXw'><strong id='rIDUXw'></strong><small id='rIDUXw'></small><button id='rIDUXw'></button><li id='rIDUXw'><noscript id='rIDUXw'><big id='rIDUXw'></big><dt id='rIDUXw'></dt></noscript></li></tr><ol id='rIDUXw'><option id='rIDUXw'><table id='rIDUXw'><blockquote id='rIDUXw'><tbody id='rIDUXw'></tbody></blockquote></table></option></ol><u id='rIDUXw'></u><kbd id='rIDUXw'><kbd id='rIDUXw'></kbd></kbd>

    <code id='rIDUXw'><strong id='rIDUXw'></strong></code>

    <fieldset id='rIDUXw'></fieldset>
          <span id='rIDUXw'></span>

              <ins id='rIDUXw'></ins>
              <acronym id='rIDUXw'><em id='rIDUXw'></em><td id='rIDUXw'><div id='rIDUXw'></div></td></acronym><address id='rIDUXw'><big id='rIDUXw'><big id='rIDUXw'></big><legend id='rIDUXw'></legend></big></address>

              <i id='rIDUXw'><div id='rIDUXw'><ins id='rIDUXw'></ins></div></i>
              <i id='rIDUXw'></i>
            1. <dl id='rIDUXw'></dl>
              1. <blockquote id='rIDUXw'><q id='rIDUXw'><noscript id='rIDUXw'></noscript><dt id='rIDUXw'></dt></q></blockquote><noframes id='rIDUXw'><i id='rIDUXw'></i>




                Mapping the Milky Way Disk Population Structures and Galactoseismology (MWDPSG): Recent progress about Stellar Parameters and Farther Outer Disk Ridge

                主 讲 人 :王海峰     博士


                地      点 :Zoom(863 5004 3883)Passcode: 253681


                We explore the performance of differentmachine learning methods for the mass and age of LAMOST sample, includingbayesian linear regression (BYS), gradient boosting decision Tree (GBDT),multilayer perceptron (MLP), multiple linear regression (MLR), random forest(RF), and support vector regression (SVR). We find that the performance ofnonlinear model is generally better than that of linear model, the GBDT and RFmethods are relatively better. The training dataset is cross matched from theLAMOST DR5 and high resolution asteroseismology data, mass and age arepredicted by random forest method or convex hull algorithm. The test datasetshows that the median relative error of the prediction model for the mass oflarge sample is 0.03 and meanwhile, the mass and age of red clump stars are0.04 and 0.07. We also compare the predicted age of red clump stars with therecent works and find that the final uncertainty of the RC sample could reach18% for age and 9% for mass, in the meantime, final precision of the mass forlarger sample with different type of stars could reach 13% , withoutconsidering systematics, all these are implying that this method could bewidely used in the future. Moreover, also with LAMOST data, we present the timetagging for the well-known ridge structures of the outer disk beyond 12 kpc. Wedetect six long-lived ridge structures, find that there might be two kinds ofdynamical origins with possible coupling mechanisms. Furthermore, thecomparison between the north and south hemispheres of the Galaxy does not showa clear asymmetry in the phase space location even though the amplitude isasymmetrical. Finally, we find that diagonal ridge structures may affect theshape of the rotation curve, which is manifested as fluctuations andundulations on top of a smooth profile.


                Hai-Feng Wang (王海峰) is now a Postdoctoral researcher of Enrico FermiResearch Center of Rome in Italy. He got his Ph.D in 2018 from NationalAstronomical Observatories of Chinese Academy of Sciences, then became thePostdoctoral researcher of Yunnan University as LAMOST Fellow in China from2018-2021, and CNRS-K.C.Wong Fellow of Paris Observatory in France from2021-2022. His interests focus on the Milky Way and Local Universe Seismology,Archaeology, Disk systems & Dark matter nature with the data, modeling andsimulations.

                发布时间:2022-06-09 10:08:21

                版权所有河北师范大◤学    冀ICP备18011017号-3  

                冀公网□安备 13010802000630号

                地址:河北省石家庄市南二环东路20号 邮编:050024 



                冀公网安备 13010802000630号