Fig. 1
From: Artificial intelligence-based nomogram for small-incision lenticule extraction
Results of RMSEs and accuracies of multiple linear regression (linear), decision tree (tree), AdaBoost, XGBoost, and MLP with hidden layer number of 1 (MLP_1), 2 (MLP_2), 4 (MLP_4), 8 (MLP_8), and 16 (MLP_16). Axis: astigmatism axis; Accuracy_0: accuracy with threshold of zero; Accuracy_25: accuracy with threshold of 0.25 D for sphere and cylinder and 25° for astigmatism axis nomograms