Machine Learning in nanoarchitectonics

Perhaps no so visible and even difficult to notice at a quick glance, the links between nanoarchitectonics and machine learning are strong and profound both historically and thematically. From ancient times through middle-ages to modern digital world, mathematics has played an important role and made an impact on many areas, including what has emerged now as nanoscience and nanoarchitectonics. In this review, we analyze artificial intelligence, machine learning and deep learning for discovery, prediction, optimization, characterization and imaging in nanoarchitectonics. Although three more general parts are highlighted: (1) atomic and molecular sciences; (2) nanotechnology for colloids and nanofilms; (3) micro- and macro- technologies, application of machine learning in nanotechnology for colloids and nanofilms (2) is particularly relevant and important, because nanofabricated structures do not coincide with projected nano-designs. In machine learning, eXplainable Artificial Intelligence (XAI) is becoming an important area helping humans to understand why a machine would make such a decision – here, it is scrutinized through analyzing interpretability, time, accuracy, parameters (ITAP) matrix. Eventually, optimization of materials design and fabrication is linked with autonomous synthesis which is discussed in perspectives finalized with conclusions, which provides the summary and inherent links between these fields.