The Ensembles of Machine Learning Methods for Survival Predicting after Kidney Transplantation

Machine learning is used to develop predictive models diagnose different diseases, particularly kidney transplant survival prediction. The paper the collected dataset of patients’ individual parameters predict critical risk factors associated with early graft rejection. Our study shows high pairwise correlation between a massive subset listed in dataset. Hence proper feature selection needed increase quality prediction model. Several methods are for selection, and results summarized using hard voting. Modeling onset events elements particular set made based on Kapplan-Meier method. Four novel ensembles machine built selected features classification task. Proposed stacking allows obtaining an accuracy, sensitivity, specifity more than 0.9. Further research will include development two-stage predictor.