Survival Mixture Model of Gamma Distribution for Modelling Heterogeneous Data

摘要

In this study survival mixture model of three components was proposed for the analysis of heterogeneous survival data. The proposed model constitutes of three components survival mixture model of the Gamma distribution. The properties of model were highlighted. Both simulated and real data were used to estimate the maximum likelihood estimators of the model by employing the Expectation Maximization (EM). Three different censoring percentages (10%, 20% and 40%) were employed in the simulated data to assess the performance of the proposed model with different censoring percentages. The comparison showed that the model performed well with the three censoring percentages. However, the estimated parameters were better with small censoring percentage. The real data were used to compare the proposed model with the pure classical parametric survival models corresponding to each component, the two and four components survival mixture models of the Gamma distributions. The Log-likelihood (LL) and the Akaike Information Criterion (AIC) values showed that the proposed model represents real data better than the pure classical survival model, the two and four components survival mixture models of the Gamma distributions. The proposed model showed that survival mixture models are flexible and maintain the features of the pure classical survival model and are better option for modelling heterogeneous survival data.