Table 5.

Classification metrics of classifiers.

Metric	Soluble yield best	Total yield best	Soluble fraction best
RCSF
CV accuracy	0.805	0.805	0.854
Holdout accuracy	0.900	0.900	0.900
CV5 precision	0.875	0.777	0.810
Holdout precision	1.000	1.000	0.833
CV5 recall	0.700	0.777	0.895
Holdout recall	0.857	0.875	1.000
CV5 F1 score	0.778	0.777	0.850
Holdout F1 score	0.923	0.933	0.909
ESF
CV accuracy	0.707	0.659	0.659
Holdout accuracy	0.800	0.700	0.700
CV5 precision	0.750	0.600	0.778
Holdout precision	1.000	0.857	1.000
CV5 recall	0.600	0.667	0.368
Holdout recall	0.714	0.750	0.400
CV5 F1 score	0.667	0.632	0.500
Holdout F1 score	0.833	0.800	0.571

CV5 corresponds to Stratified fivefold cross validation.