In the current day, appropriate characteristics are dogged from medical datasets by effectively allowing for meta-heuristic and machine learning systems. In addition, these methods can enable an automated, rather than manual, feature selection procedure. It is also observed that by combining the feature selection procedure with machine learning classifiers, prediction accuracy can be greatly enhanced. Several systems for providing healthcare remotely are discussed in the literature in an effort to spread healthcare more widely. Technology like the Internet of Things (IoT), fog computing, sensors, wearable devices, cloud computing, and so on all play a role in these kinds of systems. The scientific community is currently paying a lot of attention to cloud computing, fog computing, and the Internet of Things for their potential to improve the efficiency with which a variety of diseases may be diagnosed and monitored. Numerous healthcare and diagnostic systems have been built using cloud, fog, and IoT technologies. In this study, we investigate the potential of the Internet of Things, cloud computing, and fog computing for developing a reliable and efficient stroke prediction monitoring system. An ensemble classifier is used in the proposed system to make predictions. Patients are also notified of an impending stroke infection by alerts and warnings generated by the fog layer. The simulation results demonstrated the superior accuracy of the proposed system compared to the state-of-the-art models and classifiers. Users receive the stroke infection warning message from the proposed system. The suggested system achieves a lower latency rate than cloud computing and systems without cloud and fog computing, which is used to measure the effectiveness of the alert generation process.