International Day Of Women in Statistics and Data Science

Statistical training in Argentina is characterized by quantitative rigor and methodological innovation. Postgraduate programs of excellence, like the PhD in Statistics at the Universidad Nacional de Rosario, build this foundation. This session showcases this training through four doctoral theses that propose cutting-edge solutions for high dimensionality, spatial dependence, incomplete data, and outliers.
Attendees will learn about:
Associative Mapping: Verónica Lac Prugent combines Multivariate Elastic Net models to address quantitative and qualitative responses in genetic resources, ensuring the empirical robustness of associations.
Multi-Environment Trials: Julia Angelini addresses the limitations of the Sites Regression (SREG) model by proposing an Expectation-Maximization (EM) based imputation method and robust variants to reduce the impact of outliers.
Spatial Prediction: Mariel Lovatto introduces a semiparametric spatial autoregressive model that captures spatial dependence without relying on strict parametric covariance, resulting in improved predictive performance.
Zero-Inflated Data: María José Llop introduces a robust EM-type estimation approach for the partially linear zero-inflated Poisson (PLZIP) regression model, effectively handling extreme observations in complex relationships.
This panel highlights the level of Argentine science and its capacity to generate statistical knowledge with global impact.