Generalizing Scientific Machine Learning and Differentiable Simulation Beyond Continuous Models

Explore the cutting-edge developments in scientific machine learning (SciML) and differentiable simulation in this comprehensive lecture by Chris Rackauckas at the Alan Turing Institute. Delve into the mathematical aspects of generalizing differentiable simulation beyond continuous models, examining advanced cases such as chaotic simulations, stochastic simulations, and Bayesian inverse problems. Discover how SciML is expanding to incorporate complex model forms like jump diffusions and agent-based models. Gain insights into the evolving challenges of numerical stability, implementation issues, and other mathematical intricacies arising from the adoption of differentiable programming capabilities in scientific simulations.