UTA Faculty: Victoria Chen, Chen Kan, Jay Rosenberger, Kevin Schug, Shouyi Wang
Collaborators: Caroline West, Paige Wicker, Blair Berger
Ph.D. Students: Linh Ho Manh, Srividya Sekar, Tony Wang
Funding: NSF 2021-24
Topics: Machine learning, surrogate optimization, multivariate modeling, multi-task learning
Description: The 2019 National Academies report, A Research Agenda for Transforming Separation Science, specifies in Chapter 3 that, “key fields of research that can provide insight and knowledge to advance separation science: materials synthesis, systems engineering, responses to external stimuli, instrumentation and characterization tools, and data science and analytics.” COSMOS researchers and collaborators are studying features that are appropriate for describing characteristics of chemical compounds and developing a surrogate optimization algorithm to identify laboratory settings on instrumentation for chemical analyses. Both unsupervised learning, to capture similarities between compounds, and supervising learning, to predict or classify the behaviors of compounds in various analyses, are relevant.