SOFTWARE SYSTEM FOR POPULATION PREDICTION

Abstract

Human population growth rate is an important parameter for real world planning. It is dependent on various demographic factors such as population infant mortality rate, life expectancy at birth and total fertility rate. Several livability factors such as the number of education institutes and health facilities also play an important role in shaping the population. To predict the population of Singapore these factors are taken into account and best performing algorithms based on previous works that are linear regression, K Nearest Neighbour (KNN), random forest classifier and ensemble learning techniques are used to train models. These models can then be used through an interface where users can enter either (i) the year whose population is to be predicted or (ii) the values of the factors taken. In the first case, the values of the factors are predicted for that year and then the population based on those predicted factors, whereas in the second case the population is predicted based on the input factor values directly. The interface presents the information in graphical form for better understanding