Speaker
Aman Raju
(University of Belgrade Astro Dept)
Description
Within the framework of LSST SER-SAG-S1 team’s Quasar Neural Process Python package for modeling quasar light curves (QNPy), we integrate Self-Organizing Maps (SOMs) and upgrade the pipeline with Attentive Latent Neural Processes, to catch more nuanced variability. We present the pilot results of our analysis of both models and features sampled from latent layers of neural process on LSST AGN Data Challenge, GAIA, ZTF, and Swift quasar light curves.
Primary authors
Aman Raju
(University of Belgrade Astro Dept)
Andjelka Kovacevic
(University of Belgrade-Faculty of mathematics)
Dragana Ilic
(Department of Astronomy, Faculty of Mathematics, University of Belgrade)
Ms
Iva Cvorivic-Hajdinjak
(Dept of Astronomy, Mathematics Faculty, University of Belgrade)
Luka Popovic
(Astronomical Observatory)
Marina Pavlović
