Predictive Inorganic Synthesis based on Machine Learning using Small Data sets: a case study of Hydrodynamic Diameter-controlled Cu Nanoparticles
Ensemble regression models successfully predicted hydrodynamic diameters of Cu nanoparticles using a limited dataset, achieving an R2 of 0.74 and MAE of 23.81 nm. This outperformed a classical Design of Experiments (DoE) model (R2=0.60). The study also introduced classification tasks to predict particle size categories and distribution types, showing that while small datasets limit complex models like LLMs, robust ML can guide Cu NP synthesis.