An Intelligent Framework Towards Managing Big Data in Internet of Healthcare Things

In recent years, the healthcare industry has witnessed drastic transformations with the advancements in low cost wearable sensory devices and Internet of Things (IoT). Towards this, end the convergence of Machine Learning (ML) models have facilitated promising outcomes in delivering intelligent predictive systems towards the Internet of Healthcare Things (IoHT) paradigms. A major challenge to this end remains in the service discovery among wearable technology for identifying healthcare applications. This paper explores the Genetic Algorithm (GA) approach for addressing the NP-hardness of conventional ML models for service discovery in healthcare big data. Here, three different Ml classifiers viz., LogitBoost, Naïve Bayes (NB), and Random Forest (RF) models are considered in convergence with GA approach. The performance of the algorithms is comparatively studied for different effectiveness measures such as precision, recall, F-measure, and accuracy. It was experimentally observed that the GA-RF model outperformed the predictive performance of the other classification models by providing an accuracy of 0.9413 with considerable improvement from the other two models. Thus, the proposed approach provides an intelligent low-complexity solution for service discovery in IoHT applications.