Special Colloquium

Title: Deep Neural Network in Big Astronomical Data Era

Speaker: Jian Ge (葛健教授，University of Florida)

Time: 3 pm, June 25 (Tuesday)

Location: Lecture Hall, 3rd floor

Abstract: Astronomical large-scale surveys, such as SDSS, Kepler, Gaia, LAMOST, WISE, PAN-STARRS, and ZTF, have accumulated a massive volume of data which provide an enormous challenge in data handling, processing, and analysis using traditional methods and tools. New efficient, accurate, and fast methods and tools are urgently needed for analyzing big data for new discoveries. Deep neural network with layers of mathematical functions and algorithms, which mimic the behavior of neurons in the human brain in object recognition, has been rapidly developed in astronomy and successfully applied in detection of transit planets, identification of near-Earth objects, and absorption lines in quasar spectra and is becoming a drive force in data mining and analysis in the big data era. I will provide a brief review of the state of this emerging field, present early science results from our deep neural network applications in quasar absorption line study and transit planet detection, and also offer future perspectives in this exploding field.