Machine Learning (ML) models in mobile applications demand significant computational resources. While distributed ML offers benefits like load distribution and privacy at the network edge, it

Machine Learning (ML) models in mobile applications demand significant computational resources. While distributed ML offers benefits like load distribution and privacy at the network edge, it paradoxically introduces challenges such as slower execution and high bandwidth consumption, hindering low-latency edge services.

This talk addresses this paradox, proposing solutions through dynamic ML model structures and compression techniques. Organized by the ITU Journal on Future and Evolving Technologies (ITU J-FET), this webinar is part of a series covering communications and networking. Explore the ITU J-FET and its webinar series.