Unsupervised identification of Floquet topological phase boundaries

Nonequilibrium topological matter has been a fruitful topic of both theoretical and experimental interest. A great variety exotic phases unavailable in static systems may emerge under nonequilibrium situations, often challenging our physical intuitions. How to locate the borders between different is an important issue facilitate characterization further understand phase transition behaviors. In this work, we develop unsupervised machine-learning protocol distinguish Floquet (periodically driven) phases, by incorporating system's dynamics within one driving period, adiabatic deformation time dimension, plus symmetry all into machine learning algorithm. Results from two rich case studies indicate that able reliably reveal intricate boundaries can hence be powerful tool discover novel afforded dimension.