The Analysis of Alternative Interval-Censored and Complete Data

Abstract

The paper considers an imputation algorithm of iterative single point quantiles for the analysis of the alternative interval-censored and complete data which is a new type of censoring data. The imputed virtual interval-censored data and right censored data are all given by a modified quantile imputation algorithm for general distributions, which is proposed in the paper for this new type censoring data. The simulation results indicate that the proposed modified quantile imputation algorithm seems to be applicable and stable for the estimation of parameters with all distributions. The case of Weibull distribution that has not a closed form for parameter estimations has been studied in details by using the proposed algorithm. Specially, for exponential distribution, we proved that the convergence of the algorithm holds under general conditions, which is an extension of the previous work. The work done in the paper can be used to solve the problems of parameter estimation even in those cases when it is impossible to obtain a complete sample under some complex circumstances. This may bring a new way to balance higher estimation precision and cost on data collections in practice.