Prediction of Heart Disease Based on Logistic Regression and Random Forest Models

Abstract

Heart disease was a hot topic of research in the field of medicine in the 20th century. Heart disease is one of the more common diseases of the circulatory system. When they progress to a certain stage, they all have an impact on heart function. Severe heart disease can lead to heart failure, which can also affect a patient's quality of life and can be life-threatening. And using medical data to find appropriate ways to predict heart disease can also be effective in improving national health care. Our experiment is about predicting whether people have heart disease. This experiment was conducted to let people know for themselves if they have the disease and to treat it quickly. Some pattern recognition methods, such as logistic regression and random forest, were used in this study. Our attempt was to find the more important features using a large amount of experimental data, and then this study used different combinations of features and the two classification techniques mentioned above to build a model for predicting heart disease, in which we tried to identify the important features through a large number of experiments to improve the accuracy of cardiovascular disease prediction. The results of the heart disease model were related to the type of chest pain experienced and the maximum heart rate achieved by those with the highest odds of developing the disease. The prevalence of heart disease was higher in middle-aged and older adults, followed by a smaller proportion of younger people and those with genetic disorders. The number of men suffering from heart disease was also higher than that of women. The results of this experiment amply demonstrate the validity of our random forest model and logistic regression model in this dataset.