PROPOSED SHRUNKEN VARIANCE ESTIMATOR FOR SURVIVAL FUNCTION

Abstract

One of the main purposes of statistics is to estimate parameters, and their reliability. Kaplan-Meier survival function is most commonly used technique for reliability. Several functions variance estimators have been proposed for this purpose, where each has its own advantages and disadvantages. To overcome some of the weaknesses of conventional methods, new variance estimators are proposed. The proposed variance estimators along with their alternative variance estimators were applied in Simulation studies. In addition, the estimators are also applied to the real data sets (Heart Transplant data set and Thallassaemia data set). The performances of the proposed variance estimators are remarkable as compared to its alternative variance estimators.