Abstract:
As one of the first topological magnetic materials, the Haldane phase serves as the ground state of the spin-1 antiferromagnetic Heisenberg model. If the system conserves SU(2) symmetry, the extensive model is called the Bilinear-Biquadratic (BLBQ) model. BLBQ has rich quantum phases such as Haldane, dimerized, critical (trimerized), and ferromagnetic phases. Using machine learning to distinguish such phases is not an easy task. In this talk, we provide a method to distinguish four quantum phases, including the topological (Haldane) phase by using three different relevant spin-spin correlations as inputs to feed them into machines.