Gravitational lensing is one of the most powerful tool in astrophysics. It can probe the matter distribution without assuming the nature (dark matter or baryonic matter) or the status (thermal or dynamical equilibrium) of the matter. In weak lensing, we usually measure the shape distortions of background galaxies, and from that we can estimate the foreground mass distribution. According to the different lens objects, either galaxy, cluster or large scale structures, we have different approaches to perform lensing analysis. In strong lens, we mainly study the inner part of galaxy or cluster. So far, one major problem in strong lens is the limit number of samples, which will be solved due to LSST or Euclid. But how to identify strong lens from millions of galaxies or QSOs is the new question. Therefore, we plan to use the machine learning to perform image classification, which can improve speed of the strong lens identification.
～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～～
本通知也可以在云台学术报告网页看到：
http://www.ynao.cas.cn/xwzx/xstg/
申请安排学术报告的同事请参见台学术报告管理办法：
http://www.ynao.cas.cn/xwzx/xstg/gzzd/
云台邀请团组：南方天文观测基地 中国科学院天体结构与演化重点实验室
邀请方联系人：董小波 0871-6733-3450
7月30日
机构概况
