sigmoidLayer
Description
A sigmoid layer applies a sigmoid function to the input such that the output is bounded in the interval (0,1).
Tip
To use the sigmoid layer for binary or multilabel classification problems, set the loss
function argument of the trainnet
to
"binary-crossentropy"
.
Creation
Description
creates a sigmoid
layer.layer
= sigmoidLayer
Properties
Examples
Algorithms
Extended Capabilities
Version History
Introduced in R2020b
See Also
trainnet
| trainingOptions
| dlnetwork
| convolution2dLayer
| tanhLayer
| softmaxLayer
| exportNetworkToSimulink
| Sigmoid Layer