International Day Of Women in Statistics and Data Science

This session brings together pioneering research that tackles the dual challenges of complexity and scalability in modern statistics and data science. The presentations feature cutting-edge developments in Bayesian hypothesis testing, discrete choice modeling, and streaming spatial data analysis over complex domains. The session begins with two complementary talks introducing highly efficient, closed-form Bayesian alternatives to classical multivariate and random effects testing frameworks. It then transitions to an innovative Bayesian best-worst choice model that utilizes Integrated Nested Laplace Approximation (INLA) for complex preference data analysis. Concluding the discussion is a state-of-the-art online nonparametric estimation of spatially varying coefficient models for streaming spatial data using dynamic bivariate penalized splines over triangulations. Together, these talks highlight the vital role of advanced computational methodologies, spanning objective Bayes, deterministic approximations, and sequential updating algorithms, in achieving rigorous inference and extracting robust insights from complex data structures across diverse scientific domains.