Validation and test set significance in LM and BR

2 visualizzazioni (ultimi 30 giorni)
Vahe Ghazikhanian
Vahe Ghazikhanian il 6 Mar 2019
When I use the NN Fitting GUI the defauklt values for Training, Validation and Test data sets are 70%, 15% and 15%. However, if I select Byesian Regularization the Error histogram only shows the results for Validation and tests sets only. Where as if I were to use LM algorithm then the error histogram shows contributions from all three sets! I would like to know why this is and why validation set results are not significant in the BR algorithm?

Accedi per rispondere a questa domanda.

Risposte (0)

Categorie

Scopri di più su Hypothesis Tests in Help Center e File Exchange

Tag

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by