Probing Cosmic Rays for Mass Composition - Using Graph Neural Networks

Presented virtually over Zoom, within the IceCube Collaboration in the Cosmic-Ray Working group, this talk discussed about the necessary and possible modifications which can be done to a previous analysis. The presentation then discussed about a new composition sensitive parameter which can possibly help IceCube Collaboration to do a primary-type analysis on air showers. The presentation further discussed about my work in Graph Neural Networks for primary type classification and the outlook for it's application in the analysis pipeline for cosmic-rays.