Proposal Description

Explainable Statistical Learning focuses on the development of statistical methods that provide transparent, interpretable, and trustworthy inference for complex data. The session will bring together complementary perspectives on interpretable statistical modeling, including Bayesian uncertainty quantification and global sensitivity analysis, statistical methods for ordinal and multivariate outcomes, and flexible semiparametric models for longitudinal and functional data. Emphasis will be placed on methodological advances that improve model interpretability while maintaining statistical rigor, with applications to health, official statistics, environmental sciences, sports analytics, and artificial intelligence. By highlighting uncertainty, interpretability, and principled statistical inference, the session will demonstrate the unique contribution of modern statistical learning to responsible data science.