This paper presents a novel method for predicting heart disease using hybrid machine learning techniques, specifically the Hybrid Random Forest with Linear Model (HRFLM). The authors aim to improve the accuracy of heart disease prediction by finding significant features through the application of machine learning techniques. The prediction model is evaluated using different combinations of features and various classification techniques, achieving an accuracy level of 88.7% with the HRFLM model. The study uses the Cleveland dataset from the UCI repository, which contains 303 patient records with 13 attributes relevant to heart disease prediction. The HRFLM method combines the strengths of Random Forest (RF) and Linear Model (LM) to enhance prediction accuracy. The paper also discusses the experimental setup, data preprocessing, feature selection, classification modeling, and performance evaluation. The results show that the proposed method outperforms existing models in terms of accuracy, precision, and error rate. The authors conclude that the HRFLM approach is effective in predicting heart disease and suggest further research to improve feature selection and prediction techniques.This paper presents a novel method for predicting heart disease using hybrid machine learning techniques, specifically the Hybrid Random Forest with Linear Model (HRFLM). The authors aim to improve the accuracy of heart disease prediction by finding significant features through the application of machine learning techniques. The prediction model is evaluated using different combinations of features and various classification techniques, achieving an accuracy level of 88.7% with the HRFLM model. The study uses the Cleveland dataset from the UCI repository, which contains 303 patient records with 13 attributes relevant to heart disease prediction. The HRFLM method combines the strengths of Random Forest (RF) and Linear Model (LM) to enhance prediction accuracy. The paper also discusses the experimental setup, data preprocessing, feature selection, classification modeling, and performance evaluation. The results show that the proposed method outperforms existing models in terms of accuracy, precision, and error rate. The authors conclude that the HRFLM approach is effective in predicting heart disease and suggest further research to improve feature selection and prediction techniques.