Author: Eric Vertina
Advisor: Professor Oren Mangoubi
Category: Graduate
Abstract: MXenes are a class of 2-D materials that possess a diverse set of properties and have a wide range of applications, including improving solar panels. Physical properties of MXenes need to be better understood and controlled in order to best harness MXenes’ potential sustainability contributions. This research uses an unsupervised VICReg-based graph representation learning model to transform input features from a graph dataset a representation space, which is then fed to downstream models to predict target properties.
UN SDGs:
SDG 7 – Affordable and Clean Energy
SDG 11 – Sustainable Cities and Communities
SDG 13 – Climate Action
Video Presentation:
Poster Presentation: