Prediction of Inflation in Indonesia Using Nonparametric Regression Approach Based on Local Polynomial Estimator
Year: 2020
Type: conference-paper
Abstract: Inflation is one of macroeconomics indicator can describe the economic development of the country.Inflation is one of the important factors, the high inflation can disturb the economy a country so it has been concerned by the government.Inflation can be caused by various factors, one of them come from the money supply.Inflation in Indonesia has a high variance so it needs a model based on high degree estimator.In this research, we predict inflation using by two approaches i.e. nonparametric (local model using local polynomial estimator) and parametric (global model).Inflation predicting can is used to prepare government policies to keep inflation at a stable.The Mean Absolute Percentage Error (MAPE) is used to know the accuracy predicted value.For nonparametric approaches, i.e. the local linear model has a MAPE as 4,933% and local quadratic model has a MAPE as 4,692%, both of them is highly accurate to predict.But, for the parametric approach (global model) has a MAPE as 29,43%.Based on MAPE, we conclude that the best model is the local quadratic model at predicting inflation.
Authors: V Fibriyani, N Chamidah
Institutions: Universitas Merdeka Pasuruan, Universitas Airlangga
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: hybrid
OpenAlex ID: W3093827770