Abstract:
Heart disease continues to be a significant global health concern, highlighting the need for precise and timely detection techniques. This project focuses on tackling the issue of predicting heart disease through the development of a predictive system based on machine learning. After analyzing a dataset containing 22 attributes and more than 253,000 records, we delved into different models such as Logistic Regression, Decision Trees, Random Forest, Support Vector Machines (SVM), K-Nearest Neighbors (KNN), and Naive Bayes. We utilized a range of data preprocessing techniques, including data cleaning, normalization, and feature engineering, to enhance the dataset for model development. After evaluating various models, one model stood out for its exceptional ability to capture intricate patterns and deliver outstanding results. The Convolutional Neural Network (CNN) demonstrated superior performance, achieving the highest accuracy, precision, recall, and ROC-AUC metrics. The system was successfully implemented into a user-friendly web application using Flask and Streamlit. It now offers real-time predictions based on user input. The project findings suggest that CNN provides a highly efficient method for predicting heart disease, showing great promise for practical use and future improvements.
