Data-driven criterion for the solid-liquid transition of two-dimensional self-propelled colloidal particles far from...

Area of Science:

• Physics
• Materials Science
• Complex Systems

Background:

• Conventional criteria for solid-liquid transitions diverge in far-from-equilibrium systems.
• Self-propelled colloidal particles exhibit complex behaviors not fully captured by existing models.

Purpose of the Study:

• To develop a data-driven criterion for identifying phase transitions in two-dimensional self-propelled colloidal particles.
• To address limitations of empirical criteria in nonequilibrium parameter regimes.

Main Methods:

• A hybrid machine learning approach combining unsupervised and supervised learning was employed.
• Analysis of extensive system configurations in the nonequilibrium parameter regime.
• Development of a generic data-driven evaluation function for empirical criteria.

Main Results:

• An explicit data-driven criterion for solid-liquid transition was established.
• A new nonequilibrium threshold for the long-time diffusion coefficient was identified.
• Performance of empirical criteria was systematically evaluated and improved.

Conclusions:

• Data-driven approaches offer a robust tool for investigating phase transitions in complex systems.
• The developed methods enhance the accuracy of predictions in far-from-equilibrium conditions.
• This work provides a generic framework applicable beyond colloidal systems.