Learning phase transitions: comparing PCA and SVM
arXiv:1905.08220
Abstract
A comparison of results from principal component analysis and support vector machine calculations is made for a variety of phase transitions in two-dimensional classical spin models.
22pages, 24 figures