Federated Learning and Edge ComputingSpeaker: Gautam Srivastava – Brandon, MB, Canada
Topic(s): Artificial Intelligence, Machine Learning, Computer Vision, Natural language processing
In recent years, mobile devices can be equipped with increasingly advanced computing capabilities, which opens up countless possibilities for meaningful applications. Traditionally, any cloud-based Machine Learning (ML) approach requires that data be centralized on a cloud-based server/data center. However, this can result in critical
issues related to unacceptable latency and communication inefficiency. To this end, we have seen the profileration of multi-access edge computing (MEC) to allow the intelligence to work closer to the edge, where data is originally generated. However,
conventional edge ML technologies still require personal data to be shared with edge servers. Recently, in light of increasing security and privacy concerns, the concept of Federated Learning (FL) has been introduced. In FL, end devices use their local data to train a local ML model required by the server. The end devices then send the local model updates to the server for aggregation, called a global model. FL can serve as an enabling technology in MEC since it enables the collaborative training of an ML model and also enables ML in mobile edge networks. However, in a large, complex mobile edge networks, FL still faces implementation challenges with regard to communicational costs, resource allocation, security, and privacy. In this talk, we begin with an introduction to the background and fundamentals of FL. We then discuss how FL can work in tandem with Edge Computing networks to acheive the goals of MEC. Finally, we
discuss some open research areas and specific open problems where attendees may be able to make an impact.
About this LectureNumber of Slides: 40
Duration: 60 minutes
Languages Available: English
Request this Lecture
To request this particular lecture, please complete this online form.
Request a Tour
To request a tour with this speaker, please complete this online form.
All requests will be sent to ACM headquarters for review.