Advancing Private Federated Learning - Insights and Innovations from Research

Explore a 29-minute conference talk on advancing private federated learning, presented by Filip Granqvist, a Machine Learning Research Engineer at Apple. Delve into the world of Private Federated Learning (PFL), an approach for collaborative machine learning model training between edge devices and a central server while maintaining data privacy. Gain insights into the exponential growth of PFL research and its practical applications in big tech companies. Learn about federated learning techniques, privacy preservation methods, and the unique challenges in PFL. Discover Apple's implementation of a privacy-preserving Federated Learning system and examine key results from real-world applications and published research. Get guidance on starting PFL research using simulations with open-source tools. Benefit from Granqvist's extensive experience as a core contributor to Apple's PFL project since 2018, covering applied ML research, software design, and architecture. Understand the importance of PFL in various domains such as language models, vision, and time series analysis, and its significance for a well-functioning society.

Advancing Private Federated Learning Insights & Innovations from Research @ Apple by Filip Granqvist