Abstract
Regularization methods always focus on the selection of variables (vs) and estimation of regression parameters. So it is relied upon (vs) Because it is difficult to identify the important variables in the model, if the number of common variables is very large, and to choose the most effective variables in the model.In this paper, we proposed a new method for selecting an ordinal model . This method is Bayesian reciprocal bridge regression for the ordinal model (BOrBridge ), We have developed a new hierarchical Bayesian regression model Bayesian reciprocal bridge for the ordinal model (BOrBridge). Which motivates us to suggest a Gibbs sample New to sample parameters from the posteriors . The performance of the proposed approach was examined through simulation studies and real data analysis.The results show that our proposed method (BOrBridge) After comparing it with the AIC and BIC Identifying the best model in standard ordinal regression works very well This also indicates the convergence of the construct-specific Gibbs samples to the posteriors distribution, was quick and the mixing was good . The research reached important conclusions, represented by the superiority of the proposed method over existing methods in selecting variables and estimating parameters.
