Heart Attack Prediction in Machine Learning Environment

. Healthcare aims to impart more personal, preventive and predictive clinical treatments to patients nowadays. Since digitized data acquisition is growing more, Artificial Intelligence (AI) applications such as Machine Learning (ML) using Data Mining (DM) are used in the areas which are misunderstood as the province of human expert’s fields. AI performs mimicry of human cognitive functions. Patient care requires interaction between clinicians, caretakers, managers and executives to decide on a clinical method (treatment, diagnosis and therapy) and a non-clinical method (technology chosen, budget, etc.). Due to advancements in deep neural networks, researchers and medical practitioners focus on developing an automated system to anticipate the occurrence of heart attacks. In the proposed research work, an extreme learning machine using AI and ML based diagnosis system using R-tool for heart attack prediction using classifiers (Decision Tree (DT) and Random Forest Algorithm (RFA)) is proposed. The objective of this research work is to find the most suitable classification algorithm for predicting the heart disease. The performance of classifiers is evaluated with performance metrics such as classification accuracy and error rate. The performance of the proposed system has been validated on 13 features and a reduced set of features. The feature reduction improved the performance of classifiers in the system performance metrics of accuracy and error rate. The proposed ML-based decision support system will assist the doctors in diagnosing the prediction of heart attack for the patients efficiently.