Donald Geman, a professor of applied mathematics and statistics, works at the foundation of widely used methods in machine vision, machine learning, and transcription-based cancer phenotyping. He is a member of Johns Hopkins’ Center for Imaging Science and its Institute for Computational Medicine. He also is a visiting professor with École Normale Supérieure de Cachan and INRIA in France.
Geman is recognized for his work in stochastic processes, image analysis, machine learning, and computational medicine. He is best known for his work on occupation densities for random functions, Markov random fields for image processing, and for introducing the Gibbs Sampler algorithm for Bayesian computation and randomized decision trees for classification.
He has made seminal contributions across multiple fields in applied mathematical sciences. His idea of randomized query selection (aka “random forests”) has become one of the most widely used classification methods in computational vision and biology, and the computational basis of Microsoft’s Kinect vision system. Geman also pioneered a “twenty questions” approach to pattern recognition that is the basis for diverse systems like road tracking and face detection. He proposed a highly novel method for predicting cancer phenotypes, including diagnosis, prognosis, and prediction of treatment response, from messenger RNA (mRNA) concentrations.
In work published in the Proceedings of the National Academy of Sciences in 2018, Geman and colleagues described a method to simplify complex biomolecular data about tumors, in principle making it easier to prescribe appropriate treatments for specific patients. The computational strategy transforms highly complex information into a simplified format that emphasizes patient-to-patient variation in the molecular signatures of cancer cells. Geman’s team found a way to greatly simplify the data on tens of thousands of molecular states by converting these data to binary labels, indicating whether a measurement falls within or beyond healthy levels.
His current projects in computational biology are driven by the objective of tailoring cancer treatment to an individual molecular profile by extracting information from gigantic amounts of data about normal functioning and abnormal perturbations in biological networks. These data are accumulated by new sequencing technologies and enable his group to learn algorithms to predict disease phenotypes, progression, and treatment response for individuals.
He is a member of the National Academy of Sciences and a Fellow of the Society for Industrial and Applied Mathematics (SIAM) and the Institute of Mathematical Statistics (IMS).
Geman earned a BA in English literature from Northern Illinois University in 1965 and a PhD in mathematics from Northwestern University in 1970. He worked for the University of Massachusetts’ Department of Mathematics and Statistics from 1970 to 2001, before joining the faculty of the Whiting School of Engineering.