What is ideal combination of training vs test data in neural network based technique in context of images. Can we have training data 10 times test data?

 Respuesta aceptada

Greg Heath
Greg Heath el 9 de Jun. de 2014

1 voto

The optimal ratio for training, validation and testing depends on the data.
The MATLAB default is 70/15/15
The default split for 10-fold cross-validation is 80/10/10
Hope this helps
Greg

Más respuestas (0)

Categorías

Más información sobre Deep Learning Toolbox en Centro de ayuda y File Exchange.

Preguntada:

A
A
el 10 de Mzo. de 2011

Comentada:

A
A
el 9 de Jun. de 2014

Community Treasure Hunt

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

Start Hunting!

Translated by