Abstract:
Cardiovascular disease (CVD) is a significant global health concern and a leading cause
of death. A machine learning odel is the best predictor for the early detection and
accurate prediction and prevention of CVD and the best prospects for the study. We
begin by creating and segmenting datasets rich in health-related characteristics
carefully. We play an essential role in preserving accuracy with data processing models
that power data transformation, feature selection, and eliminating outliers. By verifying
these steps, the dataset is carefully prepared for predictive modelling purposes. Various
models such as XGBoost, Logistic Regression, LightGBM, K-Nearest Neighbors,
Gaussian Naive Bayes, Random Forest, Decision Tree, Extra Tree, AdaBoost, Gradient
Boosting, Support Vector Machine, and CatBoost are tested before searching for CVD.
Our Analysis refers to teaching and testing accuracy to determine the model's
performance and generalization ability.
CatBoost Classifier has been recognized as an expert performer, demonstrating
exceptional test accuracy [1] and literacy when applied to unfamiliar data. However,
we analyze the essential features and provide valuable insight into factors in CVD
prediction, which positively influence the prognosis of CVD disease. If CatBoost
exhibits state-of-the-art accuracy levels, the hyperparameter tuning [2] offers more
capability, representing a promising avenue for future research efforts. In summary, this
study summarizes, augments, and improves the application of machine learning in
cardiovascular disease prediction. These results are not only useful for healthcare
practitioners and researchers but also underscore the importance of AI, data processing,
and feature selection in healthcare analytics. This work establishes a solid foundation
for future research efforts, which join the advancement of medical science and the
exploration of healthcare avenues.
