From mechanism-based to data-driven approaches in materials science

Abstract A time-honored approach in theoretical materials science revolves around the search for basic mechanisms that should incorporate key feature of phenomenon under investigation. Recent years have witnessed an explosion across areas a data-driven fueled by recent advances machine learning. Here we provide brief perspective on strengths and weaknesses mechanism based approaches context mechanics materials. We discuss literature dislocation dynamics, atomistic plasticity glasses focusing empirical discovery governing equations through artificial intelligence. conclude highlighting main open issues suggesting possible improvements future trajectories fields.