Application of Multivariate Adaptive Regression Splines (MARSplines) for Predicting Hansen Solubility Parameters Based on 1D and 2D Molecular Descriptors Computed from SMILES String

Table 4

Results of the solubility classification of 33 polymers in 85 solvents [39].

Data set	TP	TN	FP	FN
A	90.8% ± 7.2%	91.6% ± 7.0%	9.2% ± 7.2%	8.4% ± 7.0%
	(p = 1.00)	(p = 1.00)	(p = 1.00)	(p = 1.00)
B	91.1% ± 6.9%	92.4% ± 7.0%	8.9% ± 6.9%	7.6% ± 7.0%
	(p = 0.88)	(p = 0.66)	(p = 0.88)	(p = 0.66)
C	93.7% ± 5.7%	92.1% ± 6.9%	6.3% ± 5.7%	7.9% ± 6.9%
	(p = 0.08)	(p = 0.80)	(p = 0.08)	(p = 0.80)
D	93.0% ± 6.0%	92.3% ± 6.6%	7.0% ± 6.0%	7.7% ± 6.6%
	(p = 0.20)	(p = 0.70)	(p = 0.20)	(p = 0.70)

A, MARSplines (25, 2) model; B, [39]; C, [45]; D, [111].