Accuracy Analysis of Heart Disease Prediction using Logistic Regression in Comparison with the Linear Regression Algorithm

Abstract

Aim: The main objective of this research article is to employ the detection of heart disease by using the Logistic Regression (LR) classifier in comparison with the Linear regression (LR) model. Materials & Methods: The dataset used in this paper was collected from the UCI machine learning repository database. The sample size for the detection of heart disease was sample 60 (Group 1=30 and Group 2 =30) and calculation was performed utilizing G-power 0.8 with alpha and beta qualities of 0.05, 0.2 with a confidence interval of 95%. The detection of heart disease is performed by the Logistic Regression (LR) classifier with a number of samples (N=30) and Linear regression (LR) model with a number of samples (N=30). Results: The Logistic Regression (LR) classifier has a 88.68 percent higher accuracy rate when compared to the accuracy rate of the Linear regression (LR) model is 78.56 percent. The study has a significance value of p=0.025. Conclusion: Logistic Regression (LR) classifier provides better outcomes in accuracy rate when compared to Linear regression (LR) model for detection of heart disease.