A Healthcare Monitoring System for Predicting Heart Disease through Recurrent Neural Network

Abstract

Predicting with accuracy of heart disease is crucial for efficiently giving meantime treatment to cardiac patients before they suffer from heart attack. The goal is acquired through machine learning model with rich healthcare information about heart diseases. Different systems based on machine learning established newly for predicting as well as diagnosing heart disease. Because of smart framework lack, such systems are unable to handle high-dimensional datasets to be utilized in various data sources during prediction of heart disease. Research aims to incorporate a prediction model to diagnose heart disease with the aid of RNN (Recurrent Neural Network) and LSTM (Long Short-Term Memory). Subsequently, research involves optimization techniques like Artificial Flora Optimization (AFO) and Modified Artificial Flora Optimization (MAFO). At last, ensemble recurrent learning model can be trained to predict heart disease. Proposed system can be examined with data of heart disease which compares with conventional classifiers depend upon weighting techniques, feature fusion and feature selection. Proposed system acquires 97.3% accuracy that can be higher than other systems. The outcome expresses the discussed system which can be highly effective to heart disease prediction, correlation with existing modern methods.