EFFICIENT PREDICTION OF STROKE PATIENTS USING LOGISTIC REGRESSION ALGORITHM IN COMPARISON TO DECISION TREE ALGORITHM

Aim: The purpose of the project is to apply machine learning modelling methodologies to produce accurate stroke predictions in patients and to assess their effectiveness. Materials and Methods: Logistic Regression and Decision Tree Algorithms are the two groups employed in this paper. The algorithms have been implemented and tested on a dataset of over 5000 records of patients' medical and personal information. Each algorithm has gone through N = 20 iterations as a result of this. The G-power test, which estimates the statistical power of statistical tests, scores about 80%. Results: Our research found that Logistic Regression had a mean accuracy of 90.56 and Decision Tree Algorithm had a mean accuracy of 75.76. The statistically significant difference between the two approaches is 0.006, as verified by independent t-tests. Conclusion: This paper is intended to implement innovative approaches to increase the efficiency of stroke prediction algorithms and improve the accuracy of existing algorithms. It is observed from the results that the Logistic Regression Algorithm has a better performance than the Decision Tree Algorithm.