MSE Seminar Series: Materials Informatics: Moving Beyond Screening via Generative Machine Learning Models
Taylor Sparks, Associate Professor, University of Utah
Machine learning already enables the discovery of new materials by providing rapid predictions of properties to complement slower calculations and experiments. However, a persistent criticism of machine learning enabled materials discovery is that new materials are very similar, both chemically and structurally, to previously known materials. This begs the question, “Can machine learning ever learn new chemistries and families of materials that differ from those present in the training data?” In this talk, I will describe new generative machine learning approaches that can be used to generate structures that do not yet exist but are likely to. I will compare generative models, including variational autoencoders, generative adversarial networks, and diffusion models, which have become standard in machine learning for images. I will describe the unique challenges that we face when using tools of this nature to generate periodic crystalline structures and I’ll also describe how tools such as DiSCoVeR and SMACT can be used in conjunction to ensure chemically reasonable yet interesting outputs.
Dr. Taylor Sparks is an Associate Professor of Materials Science and Engineering at the University of Utah and recently completed a sabbatical at the University of Liverpool with support from the Royal Society Wolfson Visiting Fellow program. He holds a B.S. in MSE from the UofU, an M.S. in Materials from UCSB, and a Ph.D. in Applied Physics from Harvard University. He was a recipient of the NSF CAREER Award and a speaker for TEDxSaltLakeCity. When he’s not in the lab, you can find him running his podcast “Materialism,” creating materials educational content for his YouTube channel, or canyoneering with his 4 kids in southern Utah.