Loyola University Chicago

College of Arts & Sciences

Spotlight On: Gregory Matthews

Gregory Matthews

Statistical Innovations

Gregory Matthews, Director of the interdisciplinary Data Science Program in the College of Arts and Sciences at Loyola University Chicago, has been named a National Science Foundation (NSF) Grant Winner for cutting-edge work in new statistical methods.

This week’s spotlight shines on Gregory Matthews, PhD, associate professor in the Department of Mathematics and Statistics and director of the interdisciplinary Data Science Program in the College of Arts and Sciences at Loyola University Chicago, who was named a National Science Foundation (NSF) grant recipient in 2020 for research that will take place over the next three years.

Matthews’ NSF research project, “Statistical Analysis of Partially Observed Shapes in Two Dimensions,” focuses on the development of new statistical methodology, based on the ideas of multiple imputation, for the analysis of partially observed shapes. Multiple imputation is an approach to the problem of missing data in statistical packages. It allows for uncertainty about the missing data by creating plausible imputed data sets and appropriately combining results obtained from each of them. Missing data may seriously compromise inferences from randomized clinical trials, for example, if missing data are not handled appropriately. In addition, Matthews’ work has implications for machine learning and artificial intelligence.

The NSF-funded project proposes as a starting point leveraging the ideas of nonparametric, so-called “hot-deck” multiple imputation to shapes defined by unlabeled points, as opposed to shapes defined by landmarks. In the past few years Matthews has focused mainly on developing techniques for the statistical analysis of partially observed shapes in collaboration with biological anthropologists. He has also worked as the statistician on several applied papers that appear in anthropology journals.

Matthews is working on developing an app for anthropologists to upload images of fossils they have in their labs. The app will perform classification and statistical analysis for them on the shapes of specific fossils, such as teeth. Matthews is eager to make the statistical analysis of shapes more accessible to people who need this type of analysis, but do not necessarily have the technical ability to do it themselves. 

“Gregory Matthews’ NSF-funded grant project is a stellar example of how our faculty utilize sophisticated research methods and draw on a variety of disciplines to solve everyday problems,” says Peter J. Schraeder, Dean of the College of Arts and Sciences at Loyola University Chicago. “His innovative, interdisciplinary work in statistics has important implications across a range of disciplines, including medicine, the physical and social sciences, and the humanities.”

Some key takeaways from Matthews’ research can be illustrated in another area of interest for Matthews: statistics in sports and specifically the Houston Astros sign-stealing scandal. In a recent project, Matthews showed that the Houston Astros significantly changed their on-field behavior when they were given information about the type of pitch being thrown to them. This work will be appearing in the journal, American Statistician.

Matthews earned a Ph.D. in statistics in 2011 from the University of Connecticut.  Prior to receiving his doctorate, with a BS in actuarial science and an MS in applied statistics from Worcester Polytechnic Institute, he spent two years in a direct marketing department building predictive models. He also served a three-year post-doctoral research fellowship at the University of Massachusetts-Amherst.  His research interests include statistical disclosure control, missing data methods, statistical genetics, and statistics in sports.

This summer, Matthews will continue to work on developing methods for the statistical analysis of partially observed shapes. To read more about Matthews’ research, visit the NSF website.