Title:Detecting topological phases in the square-octagon lattice with statistical methods

Abstract:Electronic systems living on Archimedean lattices such as kagome and square-octagon networks are presently being intensively discussed for the possible realization of topological insulating phases. Coining the most interesting electronic topological states in an unbiased way is however not straightforward due to the large parameter space of possible Hamiltonians. Here we focus on the square-octagon lattice and explore via a recently developed statistical learning method the possible topological phases that can manifest when the Fermi level of the system lies at a high-order van Hove singularity. For this purpose, we analyze large data sets of randomized tight-binding Hamiltonians labeled with the corresponding topological index and, through feature engineering, we identify the observables that are associated with non-trivial topological phases. Our analysis provides a recipe to construct tight-binding Hamiltonians for the insulating phases with Chern number $\pm $1.