Dr. Ronald Parr

Biography

Ronald Parr is a Professor in the Department of Computer Science at Duke University, having held this position since 2013. His research interests primarily revolve around machine learning, decision-making processes, and the influence of noise in data-driven models. He has contributed significantly to the understanding of how different predictive models can coexist and perform equally well on real datasets, a phenomenon he refers to as the 'Rashomon Effect.' His work often explores the trade-offs that exist within partial observability in sequential decision processes, aiming to enhance the practical utility of machine learning models. With a focus on reducing complexity while maintaining predictive accuracy, Parr's research advances the field by addressing the challenges posed by noise in data generation processes and the implications for model selection in data science.

Research Interests