| Abstract |
To provide analytics for military operations, machine learning models need to be trained on data from multiple sources at the tactical edge. Due to bandwidth and data privacy constraints in tactical coalitions, it is often impractical to transmit all the data to a central location. Distributed machine learning enables model training from decentralized datasets without sharing raw data. In this demonstration, we present a system that has the ability to train machine learning models from data distributed at multiple nodes, without sending the raw data to a central place. The system includes a protocol for information exchange to enable the distributed learning from nodes with different data distributions. We further demonstrate the effectiveness of our algorithm presented in our related long paper entitled “Distributed machine learning at resource-limited tactical edge: performance bound and control algorithm”, where we proposed a control algorithm that learns system characteristics and determines the best trade-off between local update and global parameter aggregation, to minimize the learning loss for a given resource budget. We show that the algorithm can estimate the parameters related to data distribution and resource consumption, and adapt the learning process based on these estimations in real time. The demonstration will be shown on a real system with Wi-Fi connection among multiple nodes. |