Fig. 7: Experimental feedback to ANN model. | Nature Communications

Fig. 7: Experimental feedback to ANN model.

From: Machine-learning-assisted material discovery of oxygen-rich highly porous carbon active materials for aqueous supercapacitors

Fig. 7

a Comparison of the capacitances and BET surface areas reported in this work with previously reported values. For literature data, only those with high capacitances above 100 F/g were selected and plotted. b Correlated capacitance vs experimental capacitance for the ANN model with the additional experimental data. Train RMSE = 33.0, Val RMSE = 35.6, Test RMSE = 46.2. c Capacitance vs the surface areas of micropores and mesopores when doped with 2.3% nitrogen and 20% oxygen at 5 mV/s. d Specific capacitance vs O/N content for carbon materials with Smicro = 1710 m2/g, Smeso = 1050 m2/g at 5 mV/s.

Back to article page