EoT-driven hybrid ambient assisted living framework with naive Bayes-firefly algorithm

In the current decade, ambient assisted living is attracting widespread interest due to the rapidly aging global population. The cloud-based Internet of things (IoT) healthcare systems are facing many barriers to handle the big healthcare data that IoT generates. Edge of things computing is one of the promising solutions. Accordingly, this paper proposes a hybrid ambient assisted living framework with naive Bayes-firefly algorithm (HAAL-NBFA) for monitoring elderly patients suffering from chronic diseases. This architecture exploits the current advances in the IoT by using ambient and biomedical sensors to collect the data of the elderly patient and then fuse it into context states to predict the health status of the patient in real time using context-awareness techniques. The proposed HAAL-NBFA framework proposes a five-phase classification technique to handle big imbalanced datasets resulting from long-term monitoring of elderly patients. In this paper, the firefly algorithm (FA) has been used to optimize naive Bayes classifier (NB) which selects the minimum features that give the highest accuracy. The proposed NB-FA acts as a safe-fail module that decides when to stop the system and when to permit its continuation in case of sensor's failure. The experimental results proved that the proposed HAAL-NBFA had achieved high accuracy and sensitivity in predicting the health status of patients suffering from blood pressure (BP) disorders. Furthermore, the importance of NB-FA in accelerating classifications and maintaining the continuity of HAAL-NBFA's operation has been proved by experimental results.