VARIABLE SELECTION IN BAYESIAN PENALIZED LASSO LEFT CENSORED REGRESSION MODEL

A new Bayesian lasso left censored regression (NBLLCR) method is proposed. This proposed method is presented by continuous uniform distribution (-σ^2/λ,σ^2/λ) with standard exponential distribution for a mixed representation of the Laplace distribution. The proposed method is compared with several existing Bayesian and non-Bayesian method using simulation examples and real data analysis. The results of the simulation studies and real data analysis show that the proposed method perform very well compared with other approaches.