ML in Time-domain Astronomy and Astrophysics

• Jakob Nordin (HU Berlin)

Astronomical observatories map the variable sky at ever increasing frequency and depth. The resulting real-time data streams allow a wide range of scientific programs, probing particle physics, astronomy and cosmology. Efficient statistical tools, such as ML, are critical in order to effectively use these information flows. I will introduce three areas where ML models have or will become a necessary and natural pipeline component: optical artifact rejection, transient classification and FRB detection.