TERMIUM Plus^®

Record 1,  Textual support,  English

Record number: 1, Textual support number: 1 DEF

... a collection of data points that the model hasn’t seen during its training process.  1, record 1, English, - test%20dataset

Record number: 1, Textual support number: 1 CONT

A machine learning dataset is divided into training, validation, and test sets. Machine learning typically uses these datasets to teach algorithms how to recognize patterns in the data.... The test set is the final collection of unknown-good data against which [one] can measure the performance and make adjustments.  2, record 1, English, - test%20dataset

Record 1,  Textual support,  French

Record 1,  Textual support,  Spanish

Record 2,  Textual support,  English

Record number: 2, Textual support number: 1 CONT

During training, the training dataset is used to train the model. Periodically, data from the validation dataset is fed through the model, and the loss is calculated ... By comparing the training and validation loss (and accuracy, or whichever other metrics are available) over time, you can see whether the model is overfitting.  3, record 2, English, - validation%20dataset

Record 2,  Textual support,  French

Record number: 2, Textual support number: 1 CONT

[Le] jeu de données de validation [est] utilisé au cours de la phase d'apprentissage pour contrôler l'évolution de l'erreur de généralisation.  5, record 2, French, - jeu%20de%20donn%C3%A9es%20de%20validation

Record 2,  Textual support,  Spanish

Record 3,  Textual support,  English

Record number: 3, Textual support number: 1 CONT

Four simple conceptual daily rainfall-runoff models are applied to a 25-basin data set. The drainage basins are all from the UK, covering a range of sizes, topographies, soils and climates. The quality of the simulation of the observed response is classically quantified by a minimized objective function. However, in this instance, model performance is judged by a range of quantitative and qualitative measures of fit, applied to both the calibration and validation periods. These include efficiency, mean annual runoff, baseflow index, the synthetic monthly and daily flow regimes, and the flow duration curve.  1, record 3, English, - baseflow%20index

Record 3,  Textual support,  French

Record number: 3, Textual support number: 1 CONT

Critère d'évaluation de modèles conceptuels simples pluie-débit au pas de temps journalier. Résumé : Quatre modèles conceptuels simples pluie-débit au pas de temps journalier ont été appliqués à de séries de données de 25 bassins versants. Ces bassins versants, de taille, topographie, sol et climat variés, sont tous situés au Royaume Uni. La qualité de la simulation des observations est quantifiée en minimisant une fonction objective classique. Cependant, dans notre exemple, la performance d'un modèle est jugée en mesurant qualitativement et quantitativement l'ajustement, à la fois sur la période de calage et sur la période de validation. Ces mesures comprennent l'efficacité, l'écoulement moyen annuel, l'indice de débit de base, les régimes synthétiques des débits mensuels et annuel et la courbe débit-durée-fréquence.  1, record 3, French, - indice%20de%20d%C3%A9bit%20de%20base

Record 3,  Textual support,  Spanish