自定义层

为深度学习定义自定义层

对于大多数任务，您可以使用内置层。如果没有您的任务所需的内置层，则可以定义您自己的自定义层。您可以使用自定义输出层指定自定义损失函数，并定义具有可学习参数和状态参数的自定义层。定义自定义层后，您可以检查该层是否有效，是否与 GPU 兼容，以及是否输出正确定义的梯度。要查看支持的层的列表，请参阅深度学习层列表。

函数

全部展开

自定义层

`functionLayer`	Function layer (自 R2021b 起)
`checkLayer`	Check validity of custom or function layer
`setLearnRateFactor`	Set learn rate factor of layer learnable parameter
`setL2Factor`	Set L₂ regularization factor of layer learnable parameter
`getLearnRateFactor`	Get learn rate factor of layer learnable parameter
`getL2Factor`	Get L₂ regularization factor of layer learnable parameter
`networkDataLayout`	Deep learning network data layout for learnable parameter initialization (自 R2022b 起)

网络整合

`findPlaceholderLayers`	Find placeholder layers in network architecture imported from Keras or ONNX
`replaceLayer`	Replace layer in neural network
`assembleNetwork`	(Not recommended) Assemble deep learning network from pretrained layers
`PlaceholderLayer`	Layer replacing an unsupported Keras or ONNX layer

主题

自定义层概述

定义自定义深度学习层
了解如何定义自定义深度学习层。
Define Custom Deep Learning Output Layers
Learn how to define custom deep learning output layers.
Check Custom Layer Validity
Learn how to check the validity of custom deep learning layers.
基于预训练的 Keras 层组合网络
此示例说明如何从预训练的 Keras 网络中导入层、用自定义层替换不支持的层，以及将各层组合成可以进行预测的网络。

自定义中间层

Define Custom Deep Learning Layer with Learnable Parameters
This example shows how to define a SReLU layer and use it in a convolutional neural network.
Define Custom Deep Learning Layer with Multiple Inputs
This example shows how to define a custom weighted addition layer and use it in a convolutional neural network.
Define Custom Deep Learning Layer with Formatted Inputs
This example shows how to define a custom layer with formatted dlarray inputs.
Define Custom Recurrent Deep Learning Layer
This example shows how to define a peephole LSTM layer and use it in a neural network.
Specify Custom Layer Backward Function
This example shows how to define a SReLU layer and specify a custom backward function.
Custom Layer Function Acceleration
Accelerate custom layer forward and predict functions by caching and reusing traces.
Define Custom Deep Learning Layer for Code Generation
This example shows how to define a SReLU layer that supports code generation.

自定义输出层

Define Custom Classification Output Layer
This example shows how to define a custom classification output layer with sum of squares error (SSE) loss and use it in a convolutional neural network.
Define Custom Regression Output Layer
This example shows how to define a custom regression output layer with mean absolute error (MAE) loss and use it in a convolutional neural network.
Specify Custom Output Layer Backward Loss Function
This example shows how to define a custom classification output layer with sum of squares error (SSE) loss and specify a custom backward loss function.

网络合成和嵌套层

Deep Learning Network Composition
Define custom layers that contain neural networks.
Define Nested Deep Learning Layer Using Network Composition
This example shows how to define a nested custom deep learning layer.
Train Network with Custom Nested Layers
This example shows how to create and train a network with nested layers defined using network composition.

自定义层

函数

自定义层

网络整合

主题

自定义层概述

自定义中间层

自定义输出层

网络合成和嵌套层

WeChat