An Enhanced Approach to Identify and Classify Heart Disease using Machine Learning Models

Cardiovascular Disease (CVD) is the leading cause of death worldwide, expected to account for 19.05 million deaths by 2020. Heart disease is more common in younger adults because of lifestyle factors such stress, poor diet, and inactivity. To improve human health, researchers are trying to find the early causes of heart disease. In the medical field, early diagnosis lowers mortality and enhances treatment outcomes. Classification tasks related to heart disease have a great deal of potential to be improved by machine learning (ML). In the medical domain, machine learning algorithms are becoming increasingly important, particularly when using medical databases to diagnose illnesses. Accurate diagnosis of heart illness is made possible by the sophisticated data processing and analysis methods, which are also utilized to predict other ailments. The effectiveness of the KNN, Naïve Bayes, and support vector machine (SVM) approaches is compared in this study both with and without the elephant herd optimization strategy. The most popular "Cleveland Clinic Heart Disease dataset" (including 303 cases and 14 characteristics) has been taken from the Kaggle repository and used in the proposed project. The algorithm selects seven of the most relevant features; its performance was improved via the Elephant Herd Optimization approach. Out of all these classification methods, the support vector machine produced the most accurate result, with an accuracy rate of 88.59%.