Bayesian Reciprocal LASSO Composite Quantile Regression for Robust Clinical Risk Modeling

Ali Abdulmohsin Abdulraeem Al-rubaye; Ameer Musa Imran Alhseeni

Ali Abdulmohsin Abdulraeem Al-rubaye Planning Department, General Directorate of Qadisiyah Education, Ministry of Education
Ameer Musa Imran Alhseeni Planning Department, General Directorate of Qadisiyah Education, Ministry of Education

Keywords: Bayesian Quantile Regression, Reciprocal LASSO, Composite Likelihood, Outlier Robustness, Variable Selection

Abstract

Clinical data often contain outliers and irrelevant predictors that can distort inference and reduce the reliability of traditional regression methods. To address this issue, we propose a robust Bayesian variable selection framework by integrating composite quantile regression with a reciprocal LASSO prior. The method accommodates heavy-tailed errors and performs simultaneous coefficient estimation and sparsity enforcement.We evaluate the proposed model through extensive simulation studies under contamination scenarios and compare it with classical and Bayesian LASSO-based quantile regression methods. The model is further applied to systolic blood pressure data from the NHANES 2017–2018 survey to identify key lifestyle and health-related predictors. Results show that the proposed method outperforms competing approaches in terms of predictive accuracy, robustness to outliers, and variable selection stability.

References

Alhamzawi, R. (2020). Bayesian adaptive Lasso quantile regression. Communications in Statistics - Simulation and Computation, 49(6), 1526–1538.

Enas, A. A., & Mohamed, S. H. R. (2025). Bayesian skewed-t multivariate censored quantile regression for functional neuroimaging data. Central Asian Journal of Mathematical Theory and Computer Sciences, 6(3).

Koenker, R., & Bassett, G. (1978). Regression quantiles. Econometrica, 46(1), 33–50.

Kozumi, H., & Kobayashi, G. (2011). Gibbs sampling methods for Bayesian quantile regression. Journal of Statistical Computation and Simulation, 81(11), 1565–1578.

Park, T., & Casella, G. (2008). The Bayesian Lasso. Journal of the American Statistical Association, 103(482), 681–686.

Raheem, S. H. (2025). Expectile regression with adaptive elastic net penalty. International Journal of Statistics and Applied Mathematics, 10(6), 101–106.

Raheem, S. H., & Mahdi, R. H. (2025). Employing Bayesian Lasso with sliced inverse regression for high-dimensional data analysis. Al-Qadisiyah Journal for Administrative and Economic Sciences, 26(4), 97–106.

Song, Q., & Liang, H. (2017). Bayesian quantile regression with reciprocal lasso. Statistics and Its Interface, 10(1), 101–112.

Sriram, K., Ramamoorthi, R. V., & Ghosh, P. (2013). Posterior consistency of Bayesian quantile regression based on the misspecified asymmetric Laplace density. Bayesian Analysis, 8(2), 479–504.

Vehtari, A., Gelman, A., & Gabry, J. (2017). Practical Bayesian model evaluation using leave-one-out cross-validation and WAIC. Statistics and Computing, 27(5), 1413–1432.

Yang, Y., Wang, H. J., & He, X. (2016). Posterior inference in Bayesian quantile regression with asymmetric Laplace likelihood. International Statistical Review, 84(3), 327–344.

Yu, K., & Moyeed, R. A. (2001). Bayesian quantile regression. Statistics & Probability Letters, 54(4), 437–447.

Zou, H., & Yuan, M. (2008). Composite quantile regression and the oracle model selection theory. Annals of Statistics, 36(3), 1108–1126.