MODEL SELECTION TECHNIQUES WITH APPLICATION TO PREDICTING THE DELIVERY WEEK OF A CURRENT PREGNANCY

Abstract

Model selection, sometimes, referred to as variable selection, is the process of selecting a subset of independent variables for use in model building. Variable selection arises when one wants to model the relationship between a variable of interest and a subset of potential explanatory variables or predictors, but there is an uncertainty about which subset to use. Identifying the best subset among many variables is one of the hardest parts of model building process. The purpose of this article is to understudy some of the widely used variable selection techniques, and to apply these techniques to predicting the delivery week of a current pregnancy based   on some anthropometric parameters. Random sample of two hundred expectant mothers were available and their anthropometric parameters of interest were extracted from their record in the gynaecological section of Lagos University Teaching Hospital (LUTH). The findings reveal that the best model for predicting the delivery week of a current pregnancy is the model with the three predictors: Height of the expectant mother, Cervical length and Human Chorionic Gonadotropin (HCG).