The economic growth recently become more important because of its implementation widely, the economic growth concept is a measure of country or regional economy valuation. The economic growth data in this research that is measured by Gross Regional Domestic Product (GRDP) are susceptible of multicollinearity. Multicollinearity become a problem in regression analysis, especially in Ordinary Least Square (OLS) because it causes the regression coefficient estimates become not efficient. One of method to overcome multicollinearity is using Least Absolute Shrinkage and Selection Operator (LASSO). LASSO is a shrinkage method to estimate regression coefficients by minimazing residual sum of squares subject to a constraint. Because of that constraint, LASSO can shrinks coefficients towards zero or set them to exactly zero so it can do variable selection too. Based on Variance Inflation Factor (VIF), there are high correlations between predictor variables, so there is multicollinearity in growth economic data of Jawa Tengah 2013 if we use OLS. In this research, LASSO shrinks eleven coefficients estimator of predictor variables to exactly zero, so that variables considered to have not a significant influence toward model. Keywords : LASSO, Multicollinearity, Shrinkage, Gross Regional Domestic Product (GRDP)
[1]
Robert Tibshirani,et al.
The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd Edition
,
2001,
Springer Series in Statistics.
[2]
Robert V. Brill,et al.
Applied Statistics and Probability for Engineers
,
2004,
Technometrics.
[3]
R. Tibshirani.
Regression Shrinkage and Selection via the Lasso
,
1996
.
[4]
Trevor Hastie,et al.
An Introduction to Statistical Learning
,
2013,
Springer Texts in Statistics.
[5]
D. Ruppert.
The Elements of Statistical Learning: Data Mining, Inference, and Prediction
,
2004
.
[6]
Sadono Sukirno.
Ekonomi pembangunan: proses, masalah, dan dasar kebijakan / Sukirno Sadono
,
2006
.
[7]
Y. Dewi.
OLS, LASSO dan PLS Pada data Mengandung Multikolinearitas
,
2010
.