A CANONICAL CORRELATED MULTI-AGENT REINFORCEMENT FOR E-HEALTHCARE MONITORING

Abstract

Internet of Things (IoT) is becoming more popular, sensors are used to identify the patient’s condition. Heart failure is a significant problem worldwide. It is a complicated task to forecast heart illness for a medical practitioner since it requires more contribution and understanding. Heart frequency monitoring is the fundamental computation that is crucial for heart attack prediction based on parameters like blood pressure, plasma cholesterol, and hemoglobin healthcare system, which can sense different human body parameters remotely over the Internet and then send them to an automated classification system for statistical analysis using Deep Neural Learning (DNL) techniques. The classification system is based on a DL classifier that uses IoT wearable device's log dataset to predict heart diseases. To achieve more accuracy a CCMARD approach is implemented. CCMARD is a (Multi -Agent reinforcement learning) approach. It can be classified into two categories. First, the Agent based approach provides the overall theoretical verifications. Second, Formalized approach provides the theoretical results for Patient Health monitoring system. By this way, an efficient diseased patient health monitoring is carried out with minimal time consumption. For experimentation, systematic cardiovascular healthcare data is produced utilizing Kaggle dataset and medicinal gadgets to foresee the diverse patient levels of disease severity. A detailed comparative analysis is carried out and the simulation outcome ensured the goodness of the CCMARD method over the compared methods under various aspects.