An extended Maxwell semiparametric regression for censored and uncensored data

We propose a semiparametric regression defined from the generalized odd log-logistic Maxwell distribution under the cubic spline with linear and nonlinear effects for censored and uncensored data. The parameter estimates for this regression are determined using penalized likelihood. Global influence diagnostics and quantile residuals are addressed in new standards. Several Monte Carlo simulations investigate the consistency of the estimates. The usefulness of the new regression is illustrated by means of two applications to real data.

Automated summary

We propose a semiparametric regression approach based on the generalized odd log-logistic Maxwell distribution, utilizing cubic spline with linear and nonlinear effects for modeling censored and uncensored data. The parameter estimates are determined using penalized likelihood. New standards for global influence diagnostics and quantile residuals are addressed. Monte Carlo simulations are conducted to investigate the consistency of estimates. The usefulness of the new regression is demonstrated through applications to real data.