Modeling the dependence of respiration and photosynthesis upon light, acetate, carbon dioxide, nitrate and ammonium in Chlamydomonas reinhardtiiusing design of experiments and multiple regression

Table 4 Summary of cross-validation and experimental validation of the 2nd-round models

	2nd-round models				K-fold cross-validation (average values fork = 5)						Experimental validation sets		Average scale
	2nd-round models				Training models				Test sets		Experimental validation sets
	R²	R²adj.	MAE_F	RMSE_F	R²	R²adj.	MAE_F	RMSE_F	MAE_CV	RMSE_CV	MAE_EV	RMSE_EV
CR	0.81	0.78	2.2	2.9	0.81	0.79	2.2	2.8	2.6	3.1	1.9	2.2	7.4
MA _CYT	0.64	0.61	2.7	3.5	0.66	0.62	2.6	3.5	3.2	3.9	1.7	2.1	7.5
MA _ALT	0.64	0.59	2.2	3.0	0.65	0.59	2.1	3.0	2.7	3.2	1.8	1.9	6.8
ΦPSII ₈₀₀	0.66	0.65	0.053	0.068	0.65	0.64	0.053	0.068	0.054	0.067	0.068	0.078	0.201
NPQ ₈₀₀	0.48	0.41	0.062	0.084	0.50	0.41	0.061	0.083	0.076	0.093	0.039	0.044	0.205
P ₈₀₀	0.60	0.54	13.6	19.6	0.60	0.54	13.4	19.6	18.4	23.7	17.3	23.4	43.9

Cross-validation was performed by the k-fold method with k = 5; the R², R² adjusted, MAE and RMSE presented for the training models (“F”) and the test data sets (“CV”) are the average values compiling the 5 iterations. For each response, the average scale (i.e. the difference between the mean and minimal experimental values of the DOE study) is also provided as a reference to assess the importance of the deviations. R² adj., R² adjusted.

ISSN: 1752-0509