Explore how the MELLODDY initiative leverages Kubernetes to accelerate drug discovery through a collaborative machine learning platform. Learn about the groundbreaking 18.4M EUR project that unites 10 leading pharmaceutical companies and 7 tech partners to develop a federated learning system that preserves data privacy and commercial confidentiality. Discover the architecture of the platform, including the application layer, compute plans, and Kubernetes infrastructure. Gain insights into the challenges of working with sensitive and non-sensitive data, handling bad data, and developing different models across organizations. Understand the potential of this approach to enable cooperative competition in IP-sensitive industries and revolutionize the drug discovery process.
Accelerating Drug Discovery by Competitive Cooperation Through Open Source
CNCF [Cloud Native Computing Foundation] via YouTube
Overview
Syllabus
Introduction
What is Melody
Why build a federated learning platform
What is federated learning
federated learning
okin connect
Application layer
Compute plans
Architecture of Connect
Kubernetes
Contact Information
No Organization
Sensitive Unsensitive Data
Bad Data
Different Models
Taught by
CNCF [Cloud Native Computing Foundation]
