A new family of robust regression models for overdispersed count response data

SIN RESUMEN

The main objective of this research project is to propose a new family of robust regression models for overdispersed count data. We plan to study their properties and show its usefulness by applying it to real datasets.

- Propose a new family of robust regression models for overdispersed count data. - Study the theoretical properties of the new models and perform simulation studies to evaluate their performance in terms of fit, robustness and detection of outliers. - Implement the new models into a package or into a general estimation framework like GAMLSS or BAMLSS. - Apply the new models to real datasets and compare these with other traditional models for count data regression.

Investigacion basica

Disciplinario

CONCURSO ANUAL DE INVESTIGACIÓN

• 41 — Estadística

Ciencias naturales - Matemáticas - Estadísticas, Probabilidad

PONTIFICIA UNIVERSIDAD CATÓLICA DEL PERÚ