Scaling Up AI: Embedding Physics Modeling into End-to-end Learning and Harnessing Random Projection

Explore cutting-edge approaches to AI-driven scientific discovery in this hour-long webinar from Inside Livermore Lab. Delve into the challenges of learning first-principle physics models from experimental data and discover innovative solutions to overcome computational bottlenecks. Learn about an end-to-end framework that embeds large-scale physics simulations into learning processes, allowing for efficient gradient propagation. Examine how Locality Sensitive Hashing (LSH) can be harnessed to scale up both forward simulation and backward learning of first-principle models. Gain insights into the application of these techniques for studying nano-structure evolutions in materials under extreme conditions. Led by Dr. Yexiang Xue, an expert in bridging constraint-based reasoning with machine learning, this webinar offers a deep dive into the future of AI-powered scientific research and decision-making under uncertainty.

DDPS | Scaling Up AI: Embedding Physics Modeling into End-to-end Learning and Harnessing Projection