Unsupervised learning of topological phase transitions using the Calinski-Harabaz index
Jielin Wang, Wanzhou Zhang, Tian Hua, Tzu-Chieh Wei
Abstract
The authors propose a framework to deal with unsupervised machine learning of both topological phases and non-topological ones. From this, the Calinski-Harabaz index score can be used to probe phase transitions.
Topics & Concepts
Unsupervised learningArtificial intelligenceTopological data analysisComputer scienceIndex (typography)Phase (matter)MathematicsTopology (electrical circuits)Phase transitionPattern recognition (psychology)Supervised learningMachine learningStability (learning theory)Topological indexSemi-supervised learningPhysicsTerm (time)Transition (genetics)Artificial neural networkTraining setAlgorithmGroup (periodic table)Quantum many-body systemsMachine Learning in Materials ScienceTheoretical and Computational Physics