A Fair Approach to Heart Disease Prediction: Leveraging Machine Learning Model

Abstract

This study focuses on the tasks of diagnosing and predicting diseases, which are crucial, for accurately classifying and treating them by cardiologists. By utilizing the increasing use of machine learning in the field in pattern recognition from data this research introduces a specialized model that aims to predict cardiovascular diseases. The main objectives of this model are to reduce misdiagnosis rates and minimize fatalities. To achieve these goals the proposed approach combines Logistic Regression with a fairness component. The model is trained using a real world dataset consisting of 70,000 instances obtained from Kaggle. The dataset is split into 70% for training and 30% for testing purposes to evaluate accuracy and fairness metrics at values of Logistic Regression. Through reweighing techniques applied to the model improvements, in both accuracy and fairness are observed. In conclusion this research suggests that machine learning models that prioritize fairness demonstrate performance by achieving an accuracy rate of 72% with a fairness value of 0.009.