Application of Bayesian Variable Selection in Linear Regression Models Based on High-Dimensional Data

The Bayesian variable selection is the most popular approach to statistical method to deal with the high-dimensional data. Since the high-dimensional data, where the number of covariates is larger than the number of observations ( $$p>>n$$ ) is challenge. Moreover, those covariates are highly correlated with each other since each covariate is the covariates on the econometric. This research focused on the Bayesian variable selection compared to penalised approach. Although, these are many covariates in the economy content such as Consumer Price Index (CPI), Gross Domestic Product (GDP). Those covariates are linked to the dependent variable that are rarely when compared to the hundred thousands of those covariates. In this study, we use Bayesian variable selection and LASSO in linear regression model. Moreover, there are expanded for application in real dataset about Consumer Price Index (CPI) between 2015 to 2021 in Thailand. The simulation studies results and the results from the real datasets indicated that the outperformance of Bayesian variable selection when compared to the LASSO method.