Determinants of Fertility: A Neural Network Approach

Fertility is the major determinant of population growth rate. In the early 1960’s Pakistan was the pioneer country which focused on the family planning issues to control the fertility rate so that population growth would be controlled. But it has been observed that the population growth could not be controlled at the desired level yet. The main objective of this study is to explore the factors those are affecting fertility and helps in the prediction of its levels. Analysis of this study is based on the data of 2006 and 2007 (PDHS) Pakistan Demographic and Health Survey. Multilayer Perceptron Neural Network Model is used to predict the effect of Region, status of education, wealth index, current age of respondent and contraception method on fertility. In classifying the fertility we could have achieved a rate of correct classification of 70.3% in training sample and 70.1% in holdout sample. Age is the most important determinant for predicting the levels of fertility.