Multiple imputation of masked competing risks data using machine learning algorithms

The analysis of masked cause of failure data is an important area in the reliability analysis. Prior researches mostly included masking probability as a part of likelihood function to handle masked competing risks analysis which were much time-consuming, high costly and complicated computationally. To optimize time and cost and also overcome complexity of calculation, in this paper, a new two-step approach is presented which is based on multiple imputation of masked causes of failure via some machine learning algorithms. Then, in the second step, the filled-in competing risks data are analysed using standard maximum likelihood approach. For competing risks data, the superiority of the proposed method comparing with the prior ones is evaluated in ML Estimations (MLE) of Life-time parameters via several simulation studies and applying on real data. Also, sensitivity analysis for biasness versus different sample sizes is exemplified.