This 45-minutes long meeting will take place every other Fridays starting from 6 May 2022. There will be talks from staff members, postdocs, and PhD students. Any SRG news will be introduced at the beginning of the meeting. We will start having them online, using SRG Teams. From September onwards, we expect to run them in a hybrid format.
Date: 2nd July
Speaker: Blesson Varghese
Title: Edge Machine Learning – Challenges and Opportunities
This talk will present new challenges and opportunities at the interfacee of distributed systems and machine learning (ML). ML on the edge systems is premised on (pre)processing data nearer to the source where it is generated rather than extensively relying on distant hyper clouds. This premise underpins many high-value and emerging distributed applications of scientific and societal relevance. However, many existing ML algorithms are not designed to run on relatively resource constrained environments. In this context, the concept of offloading certain ML workloads to alleviate the computational burden on frugal resources by leveraging the edge will be discussed. Using federated learning as an example, the challenges of stragglers due to computational heterogeneity and adapting to changing operational conditions are addressed.
- Title: tba