Fig. 3 | npj Computational Materials

Fig. 3

From: Chemically intuited, large-scale screening of MOFs by machine learning techniques

Fig. 3

Estimated mean absolute error as a function of the sample-size percentage used for training the model. The average estimations over 100 repetitions of the computational experiment and the 2.5th and 97.5th percentiles of the performance estimation by JAD are shown. In addition, the average of the estimation on the hold-out set is shown too. The metric shown is the Mean Absolute Error. Results for both gases CO2 and H2 are shown in each column, respectively. Values are grams of absorbed gas per grams of material

Back to article page