Towards an Intelligent Nature-Inspired Optimization Framework for Managing Healthcare Big Data

The extensive growth in healthcare information acquisition and dissemination techniques have transformed the representation of modern health industries. These platforms are highly dependent on information communication technologies (ICT) and robust learning models for performing statistical and predictive analytics in decision support systems (DSS). A crucial application of such systems is for developing DSS to support self-care activities in children with physical and motor disabilities. In this view, we suggest use of particle swarm optimization (PSO)-based random forest (RF) model for improving predictive performance of learning models with lower computational complexity. It was observed from the experimental findings that the proposed PSO-RF framework, provided highest prediction accuracy of 99.3701% with an F-measure of 99.0915. Therefore, in context to healthcare big data, the proposed framework can assist in developing personalized DSS for monitoring self-care activities in children with physical and motor disorders. This facilitates in improving the quality of life and for reducing number of personnel involved, time, and cost constraints associated with clinical trials.