Jiequn Han: Solving high-dimensional partial differential equations using deep learning #ICBS2024

Since 2017, significant advancements have been made in deep learning-based numerical algorithms for solving high-dimensional partial differential equations (PDEs). These approaches show promise in overcoming the curse of dimensionality across diverse applications. This talk will review these numerical innovations, outlining a general procedural framework. Further discussions will focus on various formulations tailored for high-dimensional scientific computing applications, from the Deep BSDE method to the variational Monte Carlo method, and promising future directions.