Performance Quantile Regression and Bayesian Quantile Regression in Dealing with Non-normal Errors (Case Study on Simulated Data)

This research discusses the performance of quantile regression and Bayesian quantile regression methods. Quantile regression uses parameter estimation by maximizing the value of the likelihood function, while Bayesian quantile regression uses parameter estimation with the Bayesian concept. The Bayesian concept in question looks for solutions from the posterior distribution with Gibbs Sampling. The purpose of the study is to compare the two methods. The data used is simulated data with a total of 100 generated data. The results obtained by the Bayesian quantile regression method are superior to the indicator used MSE with the result of 1.7445. The smallest MSE value is obtained in the model that is in quantile of 0.5