Today healthcare system has a significant role to give accurate predications in less time. Heart failure is common disease and medical field has advanced technology to diagnose the issues regarding heart. Researchers are continuously working to improve the system. This paper proposes machine learning based a new model taking the symptoms of heart failure. This paper proposes the various ensemble learning methods which helps to convert weak learner into strong learners. It also depicts the correlation between the predictors and the estimator. This paper compares and evaluates performance of the ensemble learning methods which further determine the estimated mortality rate (EMR) , false mortality rate (FMR) and an accuracy of the model. This paper will be useful for further research in medical science.

Ensemble Methods, Machine Learning, Boosting Algorithms, Bagging Algorithms, Imbalanced data