Bayesian new lasso composite quantile regression with updated Hierarchical

الملخص

The quantile regression lines are infinite, but determining an informative quantile regression line is a very hard matter, Composite quantile regression has been used to treat this issue. If the number of independent variables is large, it becomes difficult to interpret the variables accurately to overcome this problem may use the lasso procedure via the Bayesian technique via using the Laplace distribution as the prior distribution. In this paper, we will introduce a new Bayesian hierarchy is a mixture of a new formulation for Laplace prior distribution and composite quantile regression. The proposed method is compared with two other two existing methods in the same field via testing the performance of two other methods via simulation studies examples, and real data scenarios.