PENERAPAN PEMODELAN STOKASTIK DENGAN METODE POLINOM NEWTON GREGORY MAJU DAN POLINOM NEWTON GREGORY MUNDUR DALAM MEMPREDIKSI JUMLAH PENDUDUK SUMATERA UTARA

Abstract

The population growth rate of an area is influenced by several factors including fertility, mortality and migration. Meanwhile, population counts are usually carried out through a population census which is carried out every 10 years. As a metropolitan area, every year North Sumatra experiences a very rapid population growth rate. The population growth rate can be modelled with a stochastic model. In stochastic modelling refers to the percentage calculations for fertility, mortality and migration in predicting for the next census year period. Meanwhile, to calculate the population, you can use numerical methods, namely, forward newton Gregory polynomial and reverse newton Gregory polynomial. The method used in this study is a literature study where the data source used is secondary data and in this case, the researcher analyzes between three census periods, namely the period 1980 to 2010 and then predicts the period of the next census year. The purpose of this study was to determine the effectiveness of the three models in predicting the population of North Sumatra in the census period. For stochastic modelling, the prediction results obtained for the year 2010 are the percentage of total birth rate or fertility reaching ± 17.67%, the mortality reaches ± 39.61% and the migration reaches ± 42.72%. For 2020, the probability that the percentage of total birth or fertility is ± 16.83%, the mortality is ± 33.85% and the migration is ± 49.36% and for 2030 the probability of the percentage of total birth or fertility is ± 16.98%. the mortality reached ± 34.28% and the migration reached ± 48.74%. And for the forward newton Gregory polynomial method and the reverse Gregory newton polynomial, the total relative error is obtained respectively while for the reverse newton Gregory polynomial method the total error value is. Referring to the error value, the advanced Newton Gregory polynomial method has better accuracy.