Prediction of Cardiac Illness using Machine Learning Algorithms with and without Feature Selection

Predicting cardiac disease is a major challenge in medicine, as it is the leading cause of death worldwide, accounting for approximately 31% of all fatalities. Early detection and treatment of cardiac disorders is difficult, especially in poor countries where diagnostic facilities and qualified medical professionals may be lacking. The diagnostic process can also be time-consuming and complex, with medical experts relying on empirical data to make a patient's diagnosis more certain.To address this challenge, the researchers aim to use various algorithms, such as LDA, RF, GBC, DT, SVM, and KNN, along with XGBoost, Voting Classifier, MLP-ANN, and SMO, as well as feature selection techniques like SFS and RFE, with and without Feature Selection. The performance of these ten algorithms will be compared using the Datasets for Cleveland, Hungary, Switzerland, and Long Beach V, as well as the Dataset Heart Statlog, with the K-fold cross-validation approach used for verification. The front-end of the model will be developed using the Flask framework in Python, and users will log in and enter the necessary data to predict the illness. The main goal of this project is to develop a novel modelling technique that builds on previous work, as well as to create a model that is user-friendly and easily implementable in real-world scenarios.