Joe Guinness

Dr. Joe Guinness serves as an Associate Professor and Director of Undergraduate Studies in the Department of Statistics and Data Science within Cornell University's College of Agriculture and Life Sciences (CALS). His research focuses on developing computationally efficient methods for analyzing large spatial-temporal datasets, with applications spanning earth sciences, environmental monitoring, epidemiology, and precision agriculture. His work bridges theoretical statistics with practical implementation through the development of the GpGp R package for Gaussian process computation.

Dr. Guinness specializes in spatial statistics and Gaussian process modeling, particularly advancing Vecchia approximations for scalable computation. His research addresses critical challenges in interpolating satellite data, modeling environmental processes, and developing statistical frameworks for large-scale datasets. Current projects include applications in climate change modeling, soil chemistry analysis, medical imaging, and wildlife disease surveillance, with emphasis on computational efficiency and accurate uncertainty quantification.

His publication record demonstrates significant contributions to scalable spatial statistics, with recent work focusing on Vecchia approximations, Gaussian process learning, and applications to earth science problems. His research shows consistent progression toward more efficient computational methods while expanding into new application domains including epidemiology (chronic wasting disease modeling) and sports science (Vaporfly shoe impact analysis).

• Cornell Atkinson Academic Venture Fund (AVF) seed grant (2021) supporting vital interdisciplinary collaborations

Dr. Guinness actively mentors doctoral students, currently advising Megan Gelsinger at Cornell while having graduated five PhD students from North Carolina State University. His research is supported by collaborative grants across multiple disciplines, including environmental science, agriculture, and public health initiatives. The GpGp R package he developed has become a standard tool for efficient Gaussian process computation in spatial statistics.

His research group develops computational frameworks for analyzing massive spatial datasets, with particular emphasis on earth science applications requiring innovative approaches to handle satellite observations, climate model output, and environmental monitoring data. Current projects integrate statistical methodology development with practical implementation for real-world environmental challenges.