Ed Hicks, business development manager for federal and artificial intelligence at Dell Technologies (NYSE: DELL), said government agencies that intend to implement AI at the edge should consider adopting federated learning to quickly glean insights from data while ensuring the security of critical data.
In an article published on Carahsoft.com, Hicks noted that federated learning could enable agencies to leverage larger datasets at a decreased bandwidth.
“Federated models require significantly less bandwidth than other models because the information isn’t being sent back to a data center for processing,” he wrote. “If an agency has a rich dataset in the cloud and a small amount of compute at the edge, it can use federated learning to train the edge device without having to move all the data from the cloud.”
Hicks cited some of the key considerations for agencies that plan to apply AI at the edge, including the need to identify the end goal of the AI or a machine learning model.
“Other key questions include how much data an agency is trying to process and how quickly it needs the results,” he added.
The Dell Technologies executive discussed how the company works to help agencies that intend to incorporate AI at the edge, such as creating a roadmap for AI implementation, providing federated learning and analytics tools to agencies and finding ways for agencies to make the most of their data.
“We also provided a validated, containerized solution that agencies can use to quickly and easily deploy a federated learning solution in a Kubernetes environment,” Hicks added.
