Explore the world of Federated Machine Learning (FML) in this 21-minute conference talk from KubeCon + CloudNativeCon Europe 2022. Dive into the fundamental principles, various manifestations, and high-level use cases of FML, while understanding its potential to extract insights from dispersed data sources without compromising privacy. Learn how to implement FML on Kubernetes using FATE, KubeFATE, and integrate it with Kubeflow using the FATE-Operator. Discover new features being developed for FATE to address real-world scenarios, and gain insights into the challenges and opportunities encountered in practical FML applications since the previous presentation at CNOS Virtual Summit China. Covering topics such as FML system design, categories, concepts, and challenges, this talk provides a comprehensive overview of FML technology and its practical implementation in cloud-native environments.
What, Why and How of Federated Machine Learning – Implementation Using FATE, KubeFATE, and FATE-Operator for Kubeflow
CNCF [Cloud Native Computing Foundation] via YouTube
Overview
Syllabus
Intro
What, Why and How of Federated Machine Learning
Federated Learning Introduction
Federated Learning - Practical Solution to Data Availability (Silos)
Federated Learning System Design (Mobile)
Federated Learning Categories
Federated Learning Concept
Federated Learning Challenges
Federated Al Technology Enabler (FATE)
KubeFATE: FATE Cluster Management
"Low-code" FL Platform
Demo
Resources
Taught by
CNCF [Cloud Native Computing Foundation]
