Three Denison data analytics students recently competed in a national competition sponsored by the American Statistical Association and the Consortium for Advancement of Undergraduate Statistics Education.
Seniors Andrew Nguyen and Minh Le, along with graduate Dong Dong ’25, won third place in the June 2025 Intermediate Statistics class project competition with their project, A Decade Of Decline: Statistical Modeling Of Bank Of America’s Physical Branch Closures.
The project was a classroom assignment for their Advanced Predictive Methods in Data Analytics class (DA 352), taught by Assistant Professor Zhe Wang. DA 352 is designed to develop students’ understanding of the cutting-edge methods and algorithms of data analytics and how they can be used to answer questions about real-world problems.
A Decade Of Decline investigates the drivers and timing of physical branch closures at a major U.S. bank from 2014 to 2024, a period marked by rising digitalization and the COVID-19 pandemic.
According to the abstract: “Using a ten-year panel dataset, they applied tree-based classification models and survival analysis to identify key predictors of closure and estimate time-to-event risk. XGBoost classification achieves high accuracy, with deposit change, household income, and region type emerging as the most important predictors. A Cox Proportional Hazards model reveals that branches in rural or low-income areas tend to close earlier, highlighting disparities in financial access. Our combined approach addresses data imbalance and multicollinearity while offering a replicable framework for risk-based branch management. These findings contribute to ongoing discussions on financial equity and inform strategies for network consolidation.”
