Quantile regression based on the skewed exponential power distribution

Bayesian quantile regression generally relies on the asymmetric Laplace distribution (ALD) as error distribution. We consider methods for Lp-quantile based skewed exponential power (SEPD). Both and frequentist estimation procedures are outlined compared with previous work SEPD. find that our proposed greatly outperform a method in terms of estimation. Further, standard regression, we perform better root mean square (RMSE). Empirical evidence statistical properties models is provided through simulation study. real data application illustrates their performance.