A1727
Title: Improving predictive model false discovery rate estimation
Authors: Yet Nguyen - Old Dominion University (United States) [presenting]
Abstract: Although predictive modeling is fundamentally driven by error minimization, interpreting the underlying relationship between predictors and the response remains a significant research priority. Recent literature regarding False Discovery Rate (FDR) estimation within predictive frameworks is examined, and a refined procedure to improve the precision of these estimations is introduced.