MODELING OF OPEN UNEMPLOYMENT RATE AND PERCENTAGE OF POOR POPULATION WITH NONPARAMETRIC REGRESSION

Abstract

Poverty is often a topic discussed and debated in various national and international forums. The problem of poverty does not seem to be finished being discussed every day. Many factors have contributed to the high poverty rate in Indonesia, and one of them is the high unemployment rate. This research is aimed at modeling of open unemployment rate and percentage of poor population with nonparametric regression. The methods used are Nadaraya Watson Estimator (NEW), Local Polynomial Estimator (LPE) and Smoothing Spline regression. In choosing the best model using the smallest Mean Square Error (MSE) value. The pattern of the relationship between unemployment and the percentage of poor population based on the scatter plot shows an unclear pattern and does not follow a parametric regression pattern. The results of the comparison of the MSE value, smoothing spline has the smaller value than NWE