Enhancing heart disease prediction using gaussian process regression and sequential analysis for optimal sampling efficiency

Abstract

Heart disease poses a significant risk to global health, and accurate prediction of this risk is critical to public health. This study leverages the Kaggle heart disease dataset, employs machine learning models to predict heart disease risk, and introduces sequence analysis to minimize sample collection while maintaining accuracy. Of the four models, linear regression, Bayesian regression, random forest, and Gaussian process regression, Gaussian process regression proved to be the most accurate. Combined with sequential analysis, we found appropriate sampling stops to reduce acquisition costs while maintaining prediction accuracy.