Professor Minge Xie is a Distinguished Professor in the Department of Statistics at Rutgers University. His research centers on the theoretical foundations of statistical inference and the development of innovative frameworks for uncertainty quantification in artificial intelligence and data science models. He is also an expert in fusion learning, robust statistics, and both parametric and nonparametric methods. Professor Xie is widely recognized for his contributions to bridging classical statistical paradigms, advancing inferential frameworks for complex data, and developing innovative methodologies for real-world applications. In the past 25 years, his research has been continuously supported by the National Science Foundation and other funding agencies.
Professor Xie values creativity and independent thinking in research. His work not only advances statistical theory but also drives practical, interdisciplinary solutions across science, technology, and industry. With over 100 publications and numerous collaborative projects, he continues to shape the landscape of modern statistics and data science.
Research Focus
Theoretical Foundations of Statistical Inference, Data Science and AI
Professor Xie addresses fundamental questions in statistical science, particularly integrating Bayesian, fiducial, and frequentist paradigms. He co-founded the BFF research community to promote research bridging these paradigms, fostering reproducible scientific learning, and developing advanced analytical tools.
He is a major contributor to the development of the confidence distribution, a distribution-based estimator in frequentist inference analogous to the Bayesian posterior, and an original developer of the repro samples method, a framework that supports statistical inference for all types of model parameters—continuous, discrete, or non-numerical—and for diverse forms of data, including images, graphs, text, and voice. His current research leverages the repro samples method and other simulation-based inference (SBI) tools to build the theoretical foundations for interpretable artificial intelligence, tackling the challenges of “black-box” machine-learning models through rigorous and principled statistical inference.
Information Fusion and Meta-Analysis
Professor Xie has developed frameworks for combining inference information from diverse sources, including meta-analysis using confidence distributions. His work unifies classical p-value-based and model-based methods, supports robust and individualized fusion learning, and addresses challenges in both big-data and small-sample contexts with real-world applications.
Estimating Equations and Parametric/Nonparametric Models
Professor Xie’s early research was on hierarchical random-effect models, generalized estimating equations, parametric and semi-parametric models, latent modelling, and spatial-temporal clustering. His methods have been widely adopted in pharmaceuticals, toxicology, and other applied fields.
Interdisciplinary Applications
Social Science and Psychology
Professor Xie is an affiliate faculty member of the Rutgers Center for Alcohol Study. He consults on Project INTEGRATE, an NIH-supported consortium analyzing interventions for alcohol misuse. His work involves advanced statistical methods to evaluate interventions and identify target subgroups.
Medical and Pharmaceutical Research
Professor Xie has collaborated extensively with medical centers, pharmaceutical companies, and clinical researchers, developing tailored statistical methods for clinical, pre-clinical, and observational studies.
Engineering and Homeland Security
As a council member of DIMACS and CCICADA, and in collaboration with colleagues from Rutgers IE and ECE departments, Professor Xie has contributed to nuclear material surveillance, networked sensor systems, and other high-impact security and engineering applications.