Aaron Garrison

Aaron Garrison is a PhD student in chemical engineering whose research seeks to develop the fundamental theory enabling predictive simulations for high-throughput discovery of materials. Specifically, Aaron’s research employs data science and machine learning-assisted computational chemistry to more efficiently explore vast numbers of chemical candidates. His work is focused on density functional theory, a powerful technique in computational chemistry for the prediction of material properties, which are invaluable for the discovery of novel catalysts and materials. Supported by a MathWorks Fellowship, Aaron will attempt to refine an existing density functional approximation recommender to generalize across a wider chemical space through the generation of more data and explore a variety of machine learning architectures and representations to capture the relevant features that influence the properties of various materials. MATLAB enables him to analyze and visualize data in customizable, easy-to-use formats that can be shared with other researchers and used collaboratively. Aaron’s work to develop new frameworks integrating first-principles modeling with cutting-edge machine-learning techniques could lead to new discovery processes enabling the exploration of millions or even billions of material candidates.