Welcome and Overview
Peter Kairouz
Peter Kairouz is a research scientist at Google, where he focuses on federated learning research and privacy-preserving technologies. Before joining Google, he was a Postdoctoral Research Fellow at Stanford University. He received his Ph.D. in electrical and computer engineering from the University of Illinois at Urbana-Champaign (UIUC).
Introduction to TensorFlow Federated
In this talk we'll walk through how to get started with TensorFlow Federated using a straightforward example that can be extended to answer many research questions.
Emily Glanz
Emily Glanz is a software engineer at Google, building the infrastructure for federated learning and beyond. Before joining Google, she received a B.S. from the University of Iowa in electrical engineering.
Federated Optimization: Best Practices and Baselines
In this talk we'll discuss how to evaluate federated optimization algorithms. We will focus on best practices and detail a suite of baseline federated optimization tasks provided by TensorFlow Federated.
Zachary Charles
Zachary Charles is a research scientist at Google, working on the theory and practice of federated optimization. Before joining Google, he received a Ph.D. from the University of Wisconsin-Madison in applied mathematics.
Personalization & Federated Learning
In this talk, we'll give an overview of personalization approaches in federated learning. We introduce Federated Reconstruction, a method for performing partially local federated learning. We then walk through an application to matrix factorization for movie recommendation.
Karan Singhal
Karan Singhal is a software engineer at Google, working on personalization and representation learning. He's interested in socially beneficial machine learning applications. Before joining Google, he received an M.S. from Stanford with a focus on artificial intelligence.
Shanshan Wu
Shanshan Wu is a software engineer at Google, working on personalization algorithms in federated learning. Before joining Google, she received a Ph.D. from the University of Texas at Austin, with a focus on large-scale machine learning
Differentially Private Federated Learning
In this talk, we will combine federated learning with differential privacy to provide stronger privacy protection. We introduce DP-FedAvg algorithm and quantile-based adaptive clipping method. We further discuss the DP-FTRL algorithm that can achieve practical privacy-utility trade offs without assuming sampling. We demonstrate the implementation and usage of these techniques in TFF.
Galen Andrew
Galen Andrew has been working on federated learning and related technologies as a research scientist on the Quirk team since 2016. Before that, he studied optimization and machine learning at UW (PhD) and Stanford (MS), with a stint at Microsoft Research in between.
Zheng Xu
Zheng Xu is a research scientist working on federated learning at Google. He got his Ph.D. in optimization and machine learning from University of Maryland, College Park. Before that, he got his master's and bachelor's degree from University of Science and Technology of China.
Federated Analytics
In this talk, we will give an overview of federated analytics and introduce an open source API in TFF for the private heavy hitters problem.
Wennan Zhu
Wennan Zhu is a research scientist at Google, working on federated analytics and differential privacy algorithms. Before joining Google, she was a Ph.D. student at Rensselaer Polytechnic Institute, focused on mechanism design and game theory.
Research and TFF Q&A
Keith Rush
Keith Rush is a software engineer, researcher and mathematician working on TensorFlow-Federated and federated learning generally. His mathematical background is primarily in real and harmonic analysis, mathematical physics and orthogonal polynomials; at Google he has picked up knowledge of compilers and programming languages, and currently his primary work is towards the backend of TFF.
Zachery Garret
Zachary Garrett is a software engineer in Google Research working on federated learning platforms and research, particularly TensorFlow Federated. At Google he has worked on large scale distributed systems, programming languages, and performance optimizations for machine learning and data analysis.