Hanlin Sun: Network Science Ising states of matter

Abstract

Network science provides powerful tools for extracting information from interacting data. Although recently the unsupervised detection of phases of matter using machine learning has raised significant interest, the full prediction power of network science has not yet been systematically explored in this context. In this talk, I will present an-depth statistical, combinatory and topological characterisation of 2D Ising snapshot networks (IsingNets) extracted from Monte Carlo simulations of the 2D Ising model at different temperatures, going across the phase transition. Applying tools of TDA and network characterisation, we show that IsingNets are highly non-random comparing with their randomised counterparts. In addition, we identify several indicators of the phase transition. This work opens new perspectives for the unsupervised characterisation of the study of phases of matter using tools of network science.