The Bayesian regularized quantile varying coefficient model

The quantile varying coefficient (VC) model can flexibly capture dynamical patterns of regression coefficients. In addition, due to the check loss function, it is robust against outliers and heavy-tailed distributions response variable, provide a more comprehensive picture modeling via exploring conditional quantiles variable. Although extensive studies have been conducted examine variable selection for high-dimensional models, Bayesian analysis has rarely developed. regularized proposed incorporate robustness data heterogeneity while accommodating non-linear interactions between effect modifier predictors. Selecting important coefficients be achieved through selection. Incorporating multivariate spike-and-slab priors further improves performance by inducing exact sparsity. Gibbs sampler derived conduct efficient posterior inference sparse VC Markov chain Monte Carlo (MCMC). merit in estimation accuracy over alternatives systematically investigated simulation under specific levels multiple errors. case study, leads identification biologically sensible markers gene-environment interaction study using NHS data.