For each of the selected periods of time, the meteorological variables used in the model were averaged in order to obtain the predominant and most representative climate conditions.
Finally, the cross-validation log-likelihood scores per model were averaged over the ten replicates and used to rank the fit to the initial data set.
MGD and MNGD values for the null model are averages of 10,000 generations of simulations.
The phenotypic records used for fitting the models were averages over the single environment phenotypes.
Another limitation in their model was average number of worker category (Eq. 6).
The performance of the detection model was averaged for 10 runs over the 10 different datasets that were created.
The average number of multipath components ( overline{L} ) generated by this interpolated CM4 UWB WBAN channel model was averaged over 360° with 1° step (or in other words, for every single body direction).
The amount of variance explained by each successive model was averaged over the three runs.
To reduce variability, multiple rounds of cross-validation are performed using different partitions and the resulting log-likelihood scores (which measure how well the test sets were predicted by the model) are averaged over the rounds.
The statistical results of modelling performance for all ANN models were: average correlation coefficient r2 = 0.95 and relative error Er = 2.05%.
