Fig. 6 | Scientific Reports

Fig. 6

From: On the interpretability of machine and deep learning techniques for predicting CBR of stabilized soil containing agro-industrial wastes

Fig. 6

Comparison of observed and predicted CBR outputs during training (blue color) and testing (orange color) phases for approach I. Solid black line = 1:1 identity (perfect prediction); dashed line in each panel = ordinary least-squares fit (y = Ax + B). Fitted parameters (training-phase) are: ANN (A = 0.90, B = + 1.76), MARS (A = 0.98, B = + 0.35), M5p-MT (A = 0.96, B = + 0.65), LWP (A = 0.89, B = + 0.99), XGBoost (A = 0.97, B = + 0.39), LSTM (A = 1.02, B = − 0.49).

Back to article page