Work in progress: Non-Linear Conjugate Gradient method on the reduced parameter space of MPS with open boundary conditions

Claudia De Lazzari
University of Trento
Mathematics

We apply the algorithm of non-linear conjugate gradient to find an approximation of the ground state for the AKLT model on the variety of MPS with open boundary conditions. We use the minimal number of parameters needed to define the variety. This number corresponds to the dimension of the variety and it is computed as in https://arxiv.org/abs/2101.03148.
In particular, the directions of descent are vectors of this dimension, and we believe this may lead to an improvement of the calculation speed.


Back to Tensor Methods and Emerging Applications to the Physical and Data Sciences Culminating Retreat