Ensemble of Semi-Supervised Models for IoT Resource Scheduling and Sharing

This paper proposes an innovative distributed self-organizing map for IoT resource scheduling and sharing. The proposed model serves each type of QoS requirement using a local self-organizing map (SOM), and then combine SOMs with adaboosting for global resource optimization and complexity control. The proposed model is applied to heterogeneous, massive, and dynamic IoT data environment and shows comparable and useful experimental outcomes.