Type: Poster
Session: Posters (Monday & Tuesday)
Author: Kyle Mo

Abstract: The Pitt-Google Broker (PGB) is a cloud-based alert distribution service that provides real-time access to and classification of transients and variable events from astronomical surveys like the Zwicky Transient Facility (ZTF), the upcoming Vera C. Rubin Legacy Survey of Space and Time (LSST), and the Roman Space Telescope High Latitude Wide Area Survey. We use simulated training data from the Dark Energy Science Collaboration’s Extended LSST Astronomical Time-Series Classification Challenge (ELAsTiCC) to train our model. Near real-time processing and classification of alerts is especially important in the context of time-sensitive science cases. Machine learning algorithms, such as the recurrent neural network used by SuperNNova, combined with cloud resources achieve classification in less than a second per object. I will discuss the state and results of the training of SuperNNova classifier for expected alert information from LSST.