Theoretical framework uncovers integration challenges for machine learning in hierarchical metamaterials, highlighting the complementary role of physics-based modelling tools.
Understanding the limits of the design space is a key aspect in optimising complex hierarchical structures and is vital for exploring and designing novel Metamaterials. Simultaneously, abundant data (mostly text, images, and location) aggregated by multinational corporations accelerated the development of machine learning and artificial intelligence technologies. Although increasingly conceptually advanced, the origins of machine learning can be traced back to traditional statistical methods and datacentric analysis. These techniques have been used in fields where establishing relationships and using differential equations or closed-form descriptions have been challenging due to the systems’ complexity. However, well-established and validated physics-based modelling tools offer direct solutions for various physical domains relevant to metamaterials. What is the right place for the emerging machine learning techniques in that context?
No takes yet. Share an insight, caveat, or question.
Earnshaw et al. (2024) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: