Fig. 5: Experimental database and ML models facilitate inverse design.
From: A robotic platform for the synthesis of colloidal nanocrystals
a, Graphical illustration of the database for synthesizing gold NCs: O, S, D, T and I represent the orthogonal, single-factor, double-factor, triple-factor and inverse design experiments, respectively. b, Graphical illustration of the database for synthesizing double-perovskite NCs. c, ML normalized RGB–AR model for gold NCs. R, G, B represent red, green and blue, respectively. d, ML normalized RGB–size model for double-perovskite NCs. R, G, B represent red, green and blue, respectively. e, The ML-predicted correlations between SDAs (as inputs) and the AR or LSPR (as output) were identified for inverse design of targeted gold NCs. f, The ML-predicted correlations between SDAs (as inputs) and size (as output) were identified for inverse design of the double-perovskite NCs from microcrystals.
