Statistical analysis demonstrates an explicitly computable regression estimator for right-censored data with unknown errors, indicating robust large-sample consistency.
Key Points
To formulate an explicit and easily computable parameter vector estimator for linear regression models when data are randomly right-censored and the error distribution is unknown.
Derived a new mathematical estimator for parameter vectors under random right-censoring.
Identified theoretical sufficiency conditions to evaluate asymptotic statistical properties alongside a numerical example.
Demonstrated that the proposed estimator has a straightforward, closed-form definition that allows for direct computation.
Proved that the estimator is mean square consistent and asymptotically normal under specified conditions without requiring knowledge of the error distribution.