Handling uncertainty in IoT design: An approach of statistical machine learning with distributed second-order optimization

Abstract The emerging applications of Internet-of-Things (IoT) users structural and dynamic configuration of IoT devices. The designs need more concrete and generic approaches as there are variations in heterogeneity of devices, data management, and privacy protection for different IoT combinations. The objective is to yield a generic set of designs and libraries of data structure toward design methodologies. The features of sensing capacity of sensor, preciseness, persistence, and data acquisition can be used as a core statistical model and thus an emerging machine learning model is proposed to solicit a generic design methodology for IoT design. However, certain use cases are demonstrated to support the efficacy of the proposed model and in the process optimization of resources can also be maintained.