Supervised study of Novel Random Forest Algorithm for prediction of heart disease in Comparison With The Decision Tree Algorithm

Abstract

Aim: The aim of this work is to evaluate the accuracy and precision in predicting heart disease using Decision Tree (DT) and Novel Random forest (RF) Classification algorithms. Materials and Methods: Novel Random forest is appealed on a heart dataset which consists of 150 records. A framework for predicting heart disease in the medical field comparing the proposed and developed RF and DT classifiers. Sample Size Calculated as 55 in every group by using 80% G power. Sample Size Calculated using clinical analysis, with Alpha and Beta values of 0.05 and 0.5, the confidence level. confidence is 95%, nicest strength is 80% and registration rate is 1. Results: The Decision Tree classifier produces 96.42% accuracy in predicting the heart disease on the data set, whereas the Random forest classifier predicts the same at the rate of 78.45% of the time with a statistically significant difference between the two groups (p=0.004;p<0.05)with confidence interval 95%. Hence Novel Random forest is better than the Decision Tree. Conclusion: The results show that the performance of Random forest is better compared with Decision Tree in terms of both precision and accuracy.