Fig. 3: Model performance and explainable AI results. | npj Computational Materials

Fig. 3: Model performance and explainable AI results.

From: Unveiling key descriptors for electrical resistivity of alloys using high-throughput experiments and explainable AI

Fig. 3

a Prediction model performance with R2 and RMSE values. b Violin plots of feature importance determined by XAI for the total dataset. Asterisks indicate the average feature importance values. c Feature importance (IG) values of average valence electron concentration vs. electrical resistivity. d Absolute Pearson correlation coefficients for the input features and electrical resistivity (ρ) of the machine learning model.

Back to article page