Firstly, the analyser was tested for calibration and its analytical performance for BTEX monitoring compared with a more classical analytical method, namely gas chromatography coupled to a flame ionization detector (GC-FID).
The set of points (combinations tested) for calibration of the models was selected using the Dohelert network, which is a design optimization strategy that allows the effective estimation of the models coefficients.
Statistical testing for calibration has a number of drawbacks.
The performance of the models was evaluated using the C-Statistics for discrimination and the Hoshmer-Lemeshow test for calibration.
To overcome the limitations of the H-L, we combined the classic H-L test with a new statistical test for calibration, the GiViTI calibration belt [ 30, 31].
The classification task was evaluated using aROC for discrimination, and Brier Score, Brier Score Scaled, and Hosmer-Lemeshow test for calibration.
Third, we assessed the model performance of the modified CURB-65 score, which was made by incorporating CRP information into the CURB-65 model, using a calibration plot and Hosmer-Lemeshow test for calibration and AUC for discrimination.
This logistic regression model was tested for calibration with a Hosmer-Lemeshow Χ2 and the possible multicollinearity of the model was checked with an analysis of tolerance and inflation statistics.
Performances of the three systems were compared, using the area under the receiver operator characteristic curve (AUROC) to assess the discriminatory power and the Hosmer Lemeshow (HL) goodness-of-fit test for calibration.
Preliminary tests for the calibration of the corrosion process have been carried out both on bare and embedded bars.
