Authors: Subhamoy Chatterjee (SwRI), Andres Munoz-Jaramillo (SwRI), Kim Moreland (UTSA & SwRI), Maher Dayeh (SwRI & UTSA), Hazel Bain (CIRES & NOAA SWPC)

We built an ensemble of Convolutional Neural Networks that ingests both remote data and in-situ pre-flare properties, and is aimed towards prediction of true probabilities of Solar Energetic Particle event occurrence. The model-ensemble was trained with data between 1998-2013. In this poster, we show how it performs on unseen data over the period 2014-2022 and how the uncertainty achieved from the ensemble approach adds value.