Identifying and classifying quantum phases from measurable time series in many-body dynamics have significant values, yet lack details and face formidable challenges, requiring the profound knowledge of physicists. Here, to achieve a purely data-driven machine intelligent classification, we introduce a temporal fluctuation-amplified distance measure that captures the inherent temporal fluctuation complexity of dynamic evolution series in different quantum many-body phases. Significantly, the introduction of complexity-powered distance leads to remarkable improvements of unsupervised manifold learning of quantum many-body dynamics, as exemplified in models such as the discrete time crystal (DTC) and Aubry-André (AA) models. Our method does not require any prior knowledge and exhibits effectiveness even in imperfect, disordered, and noisy situations that are challenging for human scientists. Successful classification of dynamic phases in many-body systems holds the potential to enable crucial applications, including identification of tsunamis, earthquakes, catastrophes, and future trends in finance.

Close

PrevNext

View figure in article

• Quantum phase transitions
• Time crystals
• Quantum many-body systems
• Time series analysis
• Unsupervised learning