Symbolic regression with genetic programming in forecasting population growth of the Philippines / Teotima Evangelista Gorres-Abato.

Abstract : Human population is chaotic (complex) and the people in both systems are equally complex (dynamical system). Thus, the theory of dynamical system best explains the growth of human population. Symbolic regression (SR) with genetic programming (GP) is a model which uses the ideas of biological evolution to handle a complex problem in a dynamical system. Many prediction techniques were introduced and used by different researcher especially in the prediction of population. The Philippine Statistics Authority (PSA) used cohort model in predicting the population of the Philippines and uses birth rate and death rate in the national projection. This paper predicts the population of the Philippines using symbolic regression via genetic programming (GP) model and uses five demographic characteristics namely; birth rate, death rate, family planning methods adapted, life expectancy and fertility rate. For generating such model Eureqa software was used. The predicted value of population using the proposed population model was compared to the forecasted value from World Bank and PSA for the year 2010-2020.It was verified in the year 2015 when the PSA conducted the national census, and it was found out that the prediction value was much closer to the census result of 100,981,437 people.