What activation functions does patternnet use for the hidden and output layers?

4 visualizaciones (últimos 30 días)
Hi! I whant to know what activation functions patternnet uses for the hidden and output layers? The documentation says tansig, but I am not sure if it is used for both kinds of layers.

Respuesta aceptada

Gmoj Goo
Gmoj Goo el 27 de Mzo. de 2015
Patternnet uses tansig for hidden layers and softmax for output layer. You can see it using this command If you have 2 hidden layers:
net.layers{1 or 2}.transferFcn - for the hidden
net.layers{3}.transferFcn - for the output
  3 comentarios
Rahmat Izaizi Ismail
Rahmat Izaizi Ismail el 29 de Abr. de 2015
Please also advise on how to change the transfer function for the respective layer. Thank You.
Ewan McRobert
Ewan McRobert el 14 de Feb. de 2021
To change the transfer function for the first hidden layer to radial basis function use the following code:
net.layers{1}.transferFcn='radbas';
If you type the following command it will give you a complete list of valid transfer functions:
help nntransfer

Iniciar sesión para comentar.

Más respuestas (0)

Categorías

Más información sobre Deep Learning Toolbox en Help Center y File Exchange.

Community Treasure Hunt

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

Start Hunting!

Translated by