Lunch talk on Jan. 15, 2020

Strong Gravitational Lensing in the Big Data Era

Speaker:Nan Li (NAOC)

Venue:Room 2111, SWIFAR Building

Time:12:30 PM, Wednesday, January 15, 2020

Abstract:

Gravitational lensing has become one of the most powerful tools available for investigating the "dark side" of the Universe. Cosmological strong gravitational lensing, in particular, probes the properties of the dense cores of dark matter halos and offers the opportunity to study the distant Universe at flux levels and spatial resolutions otherwise unavailable. Moreover, with the capabilities of next-generation telescopes, first with LSST, and then Euclid and WFIRST, astrophysics and cosmology are stepping into the big data era, i.e., tens of billions of objects will be observed. Hence, searching and modeling strong lenses in such enormous datasets require automated approaches. For this purpose, we build programs with machine learning algorithms. In this presentation, I will first describe the construction of pipelines for automated lens-finding and -modeling adopting traditional machine learning and deep learning, then present the performance of the pipelines on simulated data created by a lensing simulation program named PICS. Furthermore, beyond machine learning, I will also introduce some attempts of analyzing strong lenses in the manner of citizen science, which is an alternative way of machine learning for solving the big data problems in astrophysics and cosmology.