how to indicate different transferFcn in a same layer when using feedforwardnet

10 次查看（过去 30 天）
锦廷 丁
锦廷 丁 2022-3-2
编辑： Krishna 2023-11-17
net = feedforwardnet([6 6 2]);
net.trainFcn='trainbr';
net.layers{1}.transferFcn='logsig';
net.layers{2}.transferFcn='tansig';
net.layers{3}.transferFcn='logsig';
net.layers{4}.transferFcn='tansig';
I know how to indicate transferFcn between two different layers.
how to indicate different transferFcn in a same layer but two different neuron

请先登录，再回答此问题。

回答（1 个）

Krishna
Krishna 2023-11-16
编辑：Krishna 2023-11-17
Hello,
Based on your question, it seems like you want to use different activation functions for different neurons within the same hidden layer. Unfortunately, this is not possible when using the 'feedforwardnet.' Each hidden layer can only have one activation function.
By setting,
net.layers{1}.transferFcn = 'tansig'
you have assigned the tan sigmoid function to the neurons in the first hidden layer and so on.
I hope this explanation clarifies the situation for you.

类别

在 Help Center 和 File Exchange 中查找有关 Deep Learning Toolbox 的更多信息

标签

产品


版本

R2021b

Community Treasure Hunt

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

Start Hunting!

Translated by