HypoSVI- Earthquake Hypocentre Inversion With Stein Variational Inference and Physics Informed Neural Networks
Offered By: Alan Turing Institute via YouTube
Course Description
Overview
Explore cutting-edge applications of physics-informed deep learning techniques in environmental and physical sciences through this 50-minute talk by Jonathan Smith from the Alan Turing Institute. Delve into three main topics: Physics Informed Neural Networks (PINNs) for solving the Eikonal Equation, probabilistic inversion in PINNs for earthquake location applications, and Neural Operators for in-ice navigation. Learn how EikoNet, a PINN-based solution, tackles the Eikonal Equation for quick travel-time calculations in geological models. Discover how the differentiability of PINNs enables the use of Stein Variational Gradient Descent for rapid earthquake location posterior distribution estimation. Gain insights into preliminary work on Neural Operators for optimizing travel-time in various sea-ice conditions to minimize fuel usage and risk in navigation.
Syllabus
Jonathan Smith - HypoSVI: Earthquake hypocentre inversion with Stein variational inference...
Taught by
Alan Turing Institute
Related Courses
Scientific ComputingUniversity of Washington via Coursera Inquiry Science Learning: Perspectives and Practices 3 - Science Content Survey
Rice University via Coursera Philosophy and the Sciences: Introduction to the Philosophy of Physical Sciences
University of Edinburgh via Coursera Natural Sciences
Modern States via Independent A mathematical way to think about biology
Udemy