Application of Robust Principal Component Analysis for Gross Regional Domestic Product of Provinces in Indonesia. Issue 1 (February 2021)

• Record Type: Journal Article
• Authors: Sanusi, Ratna Nur Mustika; Saputro, D R S; Setiyowati, R
• Is Part Of: Journal of physics. Volume 1776:Issue 1(2021)
• Publication Date: 2021-02
• Subjects: Physics -- Congresses; 530.5
• DOI: 10.1088/1742-6596/1776/1/012056
• Languages: English
• ISSNs: 1742-6588

Abstract

Abstract: Problems in the macroeconomy that are often faced by various countries are economic growth that is not on target yet. One indicator of economic growth is the value of Gross Regional Domestic Product (GRDP). The GRDP value can be predicted with a multivariate linear regression model. Multivariate linear regression will depend on the assumption of homogeneity and non-multicollinearity. In this research, a GRDP regression model was constructed based on expenditure data current prices in Indonesia and the factors that influence it. The results of the GRDP data exploration and the factors that influence are outliers data and multicollinearity occurs. So that not only a multivariate linear regression model is needed, but also a multivariate linear regression model that is robust to outliers is a robust principal component analysis (RPCA) method. This RPCA method will reduce the dimensions while overcoming the existence of outliers data. Thus, the purpose of this research is to determine the regression model with the RPCA method with robust estimation is the estimated minimum covariant determinant (MCD). The dependent variable used is the GDP of expenditure at current prices in Indonesia in 2018 and the independent variable is net export (X1 ), inventory change (X2 ), gross domestic fixed capital formation (X3 ), government consumption expenditure (X4 ), expenditure household non-profit consumption (X5 ), and consumption expenditure household (X6 ). The results show model which is constructed the GDP prediction is only affected by gross domestic fixed capital formation that is Ŷi = 0.657 + 3.068 Wi1 with R 2 = 85%