Horizontal Federating Decision Tree Learning From Data Streams: Building Intelligence in IoT Edge Networks

The paper presents Horizontal Federated Decision Tree Learning (HFDTL) system which is capable of building a collaborative model from evolving data streams in federated environment such as that of edge network where IoT gateways play the role of edge nodes processing data locally and edge server aggregates the local models using newly proposed majority-based aggregation algorithm. The popular VFDT algorithm is modified for edge nodes. The performance analysis of the HFDTL system reveals that it results in more accuracy compared to centralized VFDT for balanced partitioned data streams whereas the accuracy remains lower for unbalanced partitioned data. The communication cost also remains slightly higher for unbalanced data streams. The deployment of HFDTL system is expected to bring intelligence in IoT edge networks.