Overview
Learn how a team of Kaggle Grandmasters from NVIDIA developed a top-performing machine learning model for predicting COVID-19 mRNA vaccine stability in this 57-minute video. Explore the intricacies of the COVID-19 Vaccine Degradation Prediction Kaggle competition, where teams designed models to predict RNA degradation rates. Gain insights into mRNA vaccines, competition setup, and innovative solutions like custom dropout and neural network architectures. Benefit from the expertise of six NVIDIA data scientists and Kaggle Grandmasters as they share their approach to this critical challenge in vaccine development. Discover the potential of AI in predicting vaccine stability and enhancing the fight against COVID-19.
Syllabus
Introduction
Overview
Background
mRNA Vaccines
How a computation was organized
Competition setup
Second place solution
Custom dropout
Model architecture
Solution
Overview of Neural Networks
Neural Network Example
Questions
Next Episode
Taught by
NVIDIA Developer