NONLINEAR REGRESSION ANALYSIS OF EXPONENTIAL, VON BERTALANFFY AND DYNAMIC CARRYING CAPACITY MODELS FOR CANADA POPULATION

Abstract

This project is aimed to analyze the exponential model, Von Bertalanffy model, and the dynamic carrying capacity model based on the Canada population data. The unknown parameters in these models would be found out by using a special function in R language called ∼nls (Nonlinear Least Squares). The mathematical theory of the nls function would be introduced and explained. In order to examine if the models are appropriate for the Canada population, residual plots of the three models will be found out to compare. Furthermore, comparison method AIC (Akaike’s Information Criteria) would be used to determine the best model fitted to the Canada population among the three models. The model which shows the best result could be used as an actuarial life contingency model for the Canada population data in the future.