Table 3. Results for C₆₀ Solubility in Organic Solvents Dataset (Model Dataset 3) with Best Weighting (0.5–0–0.5) Combination as Obtained from the Retrospective Study.

model	division method	R2	Q2	MAE95%Train	QExt-F12	MAE95%Test	correct prediction in %	training set range	number of compounds outside ADa
1	sorted responseb	0.87	0.85	0.34	0.85	0.33	97.87	6.97	1 (128)
2		0.88	0.85	0.33	0.84	0.34	95.74		1 (128)
3		0.88	0.87	0.33	0.81	0.33	91.49		1 (85)
4		0.88	0.86	0.34	0.85	0.32	95.74		1 (21)
5		0.87	0.86	0.33	0.81	0.36	95.74		1 (128)
6	Kennard–Stonec	0.85	0.84	0.33	0.88	0.35	100	6.97	0
7		0.86	0.85	0.33	0.87	0.35	97.87		0
8		0.86	0.85	0.32	0.86	0.34	95.74		2 (87, 148)
9		0.79	0.77	0.37	0.84	0.39	95.74		0
10		0.86	0.85	0.33	0.87	0.34	95.74		0
11	modified-k-medoidsd	0.84	0.83	0.36	0.82	0.32	93.62	6.97	0
12		0.84	0.83	0.37	0.80	0.34	93.62		2 (64, 144)
13		0.89	0.88	0.34	0.78	0.34	89.36		0
14		0.88	0.85	0.33	0.82	0.33	95.74		1 (128)
15		0.89	0.88	0.34	0.79	0.33	89.36		0

^aAD using standardization technique is used and compound ID mentioned under parenthesis.

^bN_Training = 109, N_Test = 47.

^cN_Training = 109, N_Test = 47.

^dN_Training = 109, N_Test = 47.