High-dimensional PDEs, tensor-network, and convex optimization:

Yuehaw Khoo
University of Chicago
5747 S Ellis Ave, Suite 222

This talk presents new computational approaches for high-dimensional partial differential equations (PDEs), employing tensor networks and convex relaxations. Specifically, based on these approaches, we demonstrate the construction of inner and outer approximations to PDE solutions using low-order statistics. These in turn effectively address the curse of dimensionality.


Back to Tensor Networks