• KODURI NAGA RAJANI

Modern medicine says that heart disease is the most common health problem in the world. It is also one of the main causes of death. Heart disease is more dangerous, and if it isn't found early on, it can even have bad outcomes. To diagnose patients, methods such as electronic health records, continuous body monitoring, and figuring out their health conditions by putting medical sensor projections on their bodies and using wearable tech are used. Data mining techniques are used to efficiently sort the health data that is collected because the human body creates a huge amount of data all the time. The classification of health data is also the most important step because it needs to be done correctly to find heart disease early. Most medical experts and doctors agree that one of the main reasons diseases that can't be cured don't work is because they are hard to spot early on. Saving lives is the hardest thing for doctors to do, so it's very important to figure out what's wrong with patients right away. The goal of this study is to lower the risk of heart disease by using a good feature selection and classification-based prediction system. So, one of the main problems with the way things are done now is that heart disease can't be predicted earlier and more accurately. For that reason, this study tries to create a classifier that works well and accurately predicts the early stages of heart disease. There are two stages of research that are done in this project: feature selection and extraction. The chosen features from the combined Cleveland and Statlog heart dataset were taken out and picked using the Correlation-based Feature Selection (CFS) method. After that, the datasets were fed into both single classifiers and ensemble classifiers to see which hyper parameters were best at predicting what would happen. The results of the experiments show that optimizing hyperparameters makes the model more accurate. Compared to similar works, the proposed work had an Area under curve (AUC) of 0.997 and an accuracy of 97.92%, which was better than those works. To do this, parameter optimizations for the Rotation Forest ensemble classifier were used to get certain features from the CFS method. It was found that the results of this study were much better than those of earlier studies that focused on predicting heart disease.

Cardiovascular Disease (CVD),Correlation Feature Selection (CFS),Area Under Curve(AUC).

"A Comprehensive Survey of Heart Disease Prediction using Machine Learning Algorithms", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 6, page no.f451-f460, June-2024, Available :http://www.jetir.org/papers/JETIR2406549.pdf

"A Comprehensive Survey of Heart Disease Prediction using Machine Learning Algorithms", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 6, page no. ppf451-f460, June-2024, Available at : http://www.jetir.org/papers/JETIR2406549.pdf