Table 2 The three best parameter settings for each classifier and each data group based on the fivefold inner CV.
From: Virtual reality for assessing stereopsis performance and eye characteristics in Post-COVID
Classifier | Parameter names | Stereopsis performance | Pupil diameter | Gaze behavior | All |
|---|---|---|---|---|---|
SVM (lin kernel) | c | 0.063 (9) 0.001 (8) 0.125 (5) | 0.001 (13) 1.000 (8) 256.000 (5) | 0.250 (12) 0.001 (7) 1.000 (7) | 0.063 (9) 0.001 (8) 1.000 (6) |
SVM (rbf kernel) | c, gamma | 1.000, 0.500 (5) 4.000, 0.500 (5) 0.500, 0.125 (4) | 1.000, 2.000 (10) 1.000, 1.000 (5) 256.000, 0.125 (2) | 1.000, 4.000 (13) 0.500, 2.000 (8) 0.500, 1.000 (6) | 2.000, 1.000 (12) 1.000, 0.250 (7) 4.000, 0.250 (4) |
kNN | n_neighbors, weights | 5, uni (9) 11, uni (5) 9, uni (4) | 5, uni (12) 4, dist (6) 3, uni (3) | 4, dist (8) 3, uni (5) 6, dist (4) | 7, uni (12) 5, uni (7) 10, dist (5) |
RF | max_depth, min_samples_split, min_samples_leaf | 2, 2, 2 (11) 2, 2, 1 (10) 2, 2, 4 (10) | 2, 2, 1 (9) 2, 6, 1 (6) 2, 2, 2 (5) | 2, 2, 4 (15) 2, 2, 1 (9) 2, 2, 2 (2) | 2, 2, 1 (17) 2, 2, 2 (6) 2, 2, 4 (3) |
