LASSO Regression in Consumer Price Index Malaysia

• Main Authors: Pillaya, Khuneswari Gopal, Ravieb, Tivya, Mohd Padzil, Siti Aisyah
• Format: Conference or Workshop Item
• Language: English
• Published: 2021
• Subjects: HB221-236 Price
• Online Access: http://eprints.uthm.edu.my/2614/1/P12644_ff55d04e6fb954e2601fc15518ebb96f.pdf; http://eprints.uthm.edu.my/2614/; https://doi.org/10.1063/5.0053192
• Institution: Universiti Tun Hussein Onn Malaysia

This study is aimed to determine the factors contributing to the prediction of the total Consumer Price Index (CPI) in Malaysia through model selection using LASSO regression. The outliers are identified using the leverage values and studentized deleted residuals while the multicollinearity variables will undergo progressive elimination based on Variance Inflation Factor (VIF) values. K-fold Cross-Validation (CV) method and Mean Square Error of Prediction (MSE(P)) were used to identify the best model. Model-building without removal of outliers (Set A), model-building with the remove outliers based on leverage points and studentized deleted residuals (Set B), model-building after removal of extreme outliers based on the boxplot (Set C) were carried out. The multicollinearity variables were removed for all the three sets. The results showed that the MSE(P) of the best LASSO model in Set C is the smallest compared to the other two sets. The nine major categories such as food and non-alcoholic beverages, alcoholic beverages and tobacco, clothing and footwear, transport, communication, recreation service and culture, education, restaurants and hotels, miscellaneous goods and services have significant contribution in prediction of the total CPI in Malaysia.