Having defined a prediction rule, the next step is to evaluate its performance; and this is most often evaluated by the proportion of misclassified patients.
In the case of the FTP, we defined a prediction error as a prediction that is more/less than 50 days (± 50) from the real date of relapse onset.
Using a ΔPPrad cutoff of 13% to define a positive prediction of fluid responsiveness, a ΔVpeakbrach value of 16% or more allowed predicting with a 91% of sensitivity and 95% of specificity.
(y_{t}) together with a threshold (c) define a binary prediction recession with (s_{t}=1) whenever (y_{t} le c,) and expansion ((s_{t}=0)) whenever (y_{t}>c).
Further, we define a useful prediction as one with an RMSE smaller than the standard deviation of the experimental solubilities, to avoid being outperformed by the naïve assignment of the mean experimental solubility to all compounds [13].
We define a consensus prediction by a threshold of 0.5: if the presence fraction is greater than 0.5 the gene is present, and vice versa.
Different parameters of the joint are varied (surface preparation, adherend and bond thickness, lap length, geometrical configuration) so that it is possible to check qualitative expectations, and define a quantitative failure prediction parameter.
Once the best model was defined (OPM), a prediction score, defined by a Cox resulting linear function, was then calculated for each sample, thus defining two classes: the 'Unfavourable' class with a positive score and the 'Favourable' class with a negative score.
Furthermore, it defines a different prediction structure that includes hierarchical bidirectional pictures, outperforming traditional Group of Pictures patterns in both scenarios: single-view and multi-view.
