Programme

Session C11: New methods for data analysis: from design to model-based estimation
DescriptionThis section contains presentations that combines problems of statistical inference and estimation with issues related with the employed sampling design. The two first communications are in the setup of small area estimation, where the domain sample sizes are too small for obtaining efficient direct estimators. The third communications studies new and efficient estimation methods for multinomial logistic regression models under complex sampling designs. Finally, the last communication experiments some methods of sequential aggregation for environmental forecasting.
OrganisersLeandro Pardo (President of the Spanish Statistical Association)
ChairJean-Michel Poggi (University Paris Descartes, University Paris-Sud)
Presentation #1Multivariate Fay–Herriot models for small area estimation
Domingo Morales (Universidad Miguel Hernández de Elche)
Presentation #2Generalized coherent calibration using small area estimates
Ralf T. Münnich (Trier University)
Martin Rupp (Trier University)
Presentation #3Minimum Distance Estimators in Logistic Regression under Complex Designs
Leandro Pardo (Spanish Society of Statistics and Operations Research)
Nirian Martin Apaolaza (Universidad Complutense de Madrid)
Elena Castilla Gonzalez (Universidad Complutense de Madrid)
Presentation #4Sequential Aggregation of Heterogeneous Experts for PM10 Forecasting
Jean-Michel Poggi (University Paris Descartes, University Paris-Sud)
Benjamin Auder (University Paris-Sud)
Bruno Portier (Institut national des sciences appliquées de Rouen)
Michel Bobbia (Air Normand)

Back to programme »