Abstract:
Tensor networks methods have an impressive record of solving apparently exponentially hard problems of many-body physics. They also appear very naturally in quantum computing: a quantum circuit is essentially a tensor network. In the last few years, they have started to leave their original field to provide new algorithms for totally different problems arising in e.g. fluid dynamics, plasma physics or chemistry. A key trigger for these new developments is the advent of a new class of learning algorithms that can (potentially) map any problem onto the tensor network tool box.
In this talk, I will introduce these concepts and discuss more precisely one emerging application to quantum chemistry (how to use a many-body technique to solve a partial differential equation, how to improve on Gaussian atomic orbitals).