Skip to main content

Table 6 Validation and comparison with the unseen dataset

From: Disruption of Pseudomonas putida by high pressure homogenization: a comparison of the predictive capacity of three process models for the efficient release of arginine deiminase

Statistical parameter ADI activity (U/mL) Total protein release (g/L) Cell disruption (%)
RSM ANN SVM RSM ANN SVM RSM ANN SVM
Training dataset
R 2 0.970 0.869 0.975 0.963 0.854 0.960 0.984 0.945 0.991
MSE 0.255 1.175 0.208 0.029 0.151 0.031 2.725 9.298 1.586
RMSE 0.505 1.084 0.456 0.170 0.389 0.177 1.651 3.049 1.259
SEP 2.175 4.665 1.964 5.264 12.073 5.496 2.204 4.071 1.682
RPD 1.573 3.978 1.046 4.272 10.211 4.463 1.790 2.910 0.694
Unseen dataset
R 2 0.943 0.845 0.892 0.584 0.533 0.640 0.990 0.969 0.979
MSE 0.866 0.218 0.608 0.083 0.049 0.081 17.469 19.948 17.720
RMSE 0.931 0.467 0.780 0.288 0.222 0.285 4.180 4.466 4.209
SEP 3.628 1.821 3.039 8.116 6.264 8.053 6.873 7.344 6.922
RPD 3.469 1.523 2.509 5.966 5.154 5.703 6.728 6.478 5.384
  1. R 2 determination coefficient; MSE mean square error; RMSE root mean square error; SEP standard error of prediction; RPD relative percent deviation; A Cell mass concentration; B operating pressure; C number of passes