J. Espasandín Domínguez, C. Cadarso Suárez, G. Marra, R. Radice, F. Gude
The technological progress has led to the development of new measurement procedures in the form of functional data. In this work, we propose to incoporate this functional information within the framework of copula regression models for location, scale and shape (CGAMLSS). In this work, we propose the use of CGAMLSS with flexible additive predictors including functional effects, to model the joint distribution of two proteins that are useful in the control of individuals with diabetes. The level of glucose in patients is an important predictor of these two proteins, and in our study this is recorded every 5 minutes over several days. Therefore, glucose needs to enter the model as a functional covariate. The inclusion of glucose profiles into this type of models marks a novel contribution in diabetes research. The models can be implemented in the GJRM R package, which also includes some newly functions, aiding the biomedical researcher in the interpretation of the empirical results.
Palabras clave / Keywords: copula, diabetes, functional data, joint model, scalar-on-function regression
Programado
Sesión V05 Bioestadística
1 de junio de 2018 16:00
Sala 4
