Data-driven exploration of structure-dynamics relation in polymer systems

The latest development in hardware capabilities and computational physics have drastically increased the amount of available data while machine learning has generated many new different approaches to extract or generate information from large amounts of data. In an interdisciplinary approach to the frontiers of both polymer physics and machine learning, this project encompasses a comprehensive exploration of data-driven methods to enhance soft materials’ analysis, with a focus on meaningful descriptors, trajectories’ generation, system dynamics, interpretability, and bridging the gap between machine learning and simulation modeling approaches.