Defensa de tesis Ana Cristina Moreno Castilla "Análisis de los recién nacidos en Puerto Rico..."
ANALYSIS OF THE WEIGHT OF THE NEWBORN IN PUERTO RICO USING REGRESSION TO QUANTILES
By ANA CRISTINA MORENO CASTILLA (November 2018)
Chair: Dr. Dámaris Santana Morant
Major Department: Mathematical Sciences
Abstract: This project illustrates the use of the grouped (Smoothly Clipped Absolute Deviation) SCAD, a variable selection method based on regularization, in quantile regression to model the weight of newborns in Puerto Rico in 2009-2011. Quantile regression models allows us to study the effect of the independent variables on the different quantiles of the dependent variable and thus to have a complete idea of the relationship between these variables and the distribution of the response. Therefore, one can conclude what, prenatal care variables, sociodemographic factors, and health conditions are associated to, not only the low weight and overweight of newborns, also in any other part of the distribution of the newborn weights (for example, the median). A quantile regression model for newborn weight with several covariates was adjusted for different quanitles of interest. Some of the variables included in the model were age of the mother, weeks of gestation and sex. However, due to the large number of possible explanatory variables associated to a particular quantile of the newborn weight, a variable selection method based on regularization, namely SCAD, was implemented to come up with a subset of covariates significantly associated with the quantile of the newborn weights. Our results suggest that the factors significantly associated to the newborn weights depend on the quanitle being modeled, although there are some explanatory variables that are consistently selected regardless the quantile. For instance, the weight gain of the mother was maintained in all cases.
- Friday, November 30, 2018
- 12:30pm - 3:00pm
- GRIC - FullHub
- UPR - Recinto de Mayagüez