PhD candidate Aaron Hudson has been awarded the 2021 David P. Byar Award by the American Statistical Association’s (ASA) Biometric Section. The annual award recognizes the best Biometrics Section paper presented at the Joint Statistical Meetings by an early career investigator.
Statistical analyses are often conducted using simple models that only perform well under strong assumptions, which are often unrealistic in practice. Instead, recent interest has focused on using more flexible models that require much weaker assumptions. Hudson’s paper, “Honest Uncertainty Quantification for Infinite-Dimensional Risk Minimizers via the Restricted Gradient Test,” introduces a general approach for performing formal statistical inference when using these less restrictive models.
“My work has many relevant applications in public health and medicine,” says Hudson. “For example, my methods facilitate the use of flexible statistical models to identify risk factors for adverse health outcomes. Though I tend to think primarily about public health applications, my methods are applicable in other scientific domains as well.”
Professor Ali Shojaie, Hudson’s advisor, says the award deservedly acknowledges Hudson’s passion and dedication for conducting original and impactful research and recognizes his outstanding creativity and diverse skills.
“Aaron's paper tackles a challenging and fundamental problem in biostatistics and opens the door to new discoveries by providing an ingenious and elegant solution to an open problem in nonparametric inference that is relevant in many practical applications,” says Shojaie.
Hudson is broadly interested in developing statistical methods for efficiently analyzing complex biomedical data. In addition to the work outlined in his paper, he develops methods for analyzing biological systems to gain insight into etiology of complex diseases as well as methods for determining how a patient’s response to a medical treatment depends on patient characteristics.
— Deb Nelson