DAGNetwork

R2026b

(Not recommended) Directed acyclic graph (DAG) network for deep learning

DAGNetwork objects are not recommended. Use dlnetwork objects instead. For more information, see Version History.

To learn more about how to transition trainNetwork, SeriesNetwork, and DAGNetwork code to dlnetwork workflows, see Transition trainNetwork, SeriesNetwork, and DAGNetwork Code to dlnetwork Workflows.

Description

A DAG network is a neural network for deep learning with layers arranged as a directed acyclic graph. A DAG network can have a more complex architecture in which layers have inputs from multiple layers and outputs to multiple layers.

Creation

There are several ways to create a DAGNetwork object:

• Load a pretrained network such as squeezenet, googlenet, resnet50, resnet101, or inceptionv3. For an example, see Load SqueezeNet Network. For more information about pretrained networks, see Pretrained Deep Neural Networks.

• Train or fine-tune a network using trainNetwork.

• Import a pretrained network from TensorFlow™-Keras, TensorFlow 2, Caffe, or the ONNX™ (Open Neural Network Exchange) model format.

  • For a Keras model, use importKerasNetwork. For an example, see Import and Plot Keras Network.

  • For a TensorFlow model in the saved model format, use importTensorFlowNetwork. For an example, see Import TensorFlow Network as DAGNetwork to Classify Image.

  • For a Caffe model, use importCaffeNetwork. For an example, see Import Caffe Network.

  • For an ONNX model, use importONNXNetwork. For an example, see Import ONNX Network as DAGNetwork.

• Assemble a deep learning network from pretrained layers using the assembleNetwork function.

Note

To learn about other pretrained networks, see Pretrained Deep Neural Networks.

Properties

Layers — Network layers

Read-only: Layer array

This property is read-only.

Network layers, specified as a Layer array.

Connections — Layer connections

Read-only: table

This property is read-only.

Layer connections, specified as a table with two columns.

Each table row represents a connection in the layer graph. The first column, Source, specifies the source of each connection. The second column, Destination, specifies the destination of each connection. The connection sources and destinations are either layer names or have the form "layerName/IOName", where "IOName" is the name of the layer input or output.

Data Types: table

InputNames — Names of input layers

Read-only: cell array of character vectors

This property is read-only.

Names of the input layers, specified as a cell array of character vectors.

Data Types: cell

OutputNames — Names of output layers

Read-only: cell array

This property is read-only.

Names of the output layers, specified as a cell array of character vectors.

Data Types: cell

Object Functions

activations	(Not recommended) Compute deep learning network layer activations
classify	(Not recommended) Classify data using trained deep learning neural network
predict	(Not recommended) Predict responses using trained deep learning neural network
plot	Plot neural network architecture
predictAndUpdateState	(Not recommended) Predict responses using a trained recurrent neural network and update the network state
classifyAndUpdateState	(Not recommended) Classify data using a trained recurrent neural network and update the network state
resetState	Reset state parameters of neural network

Examples

Create Simple DAG Network

Create a simple directed acyclic graph (DAG) network for deep learning.

Train the network to classify images of digits. The simple network in this example consists of:

• A main branch with layers connected sequentially.

• A shortcut connection containing a single 1-by-1 convolutional layer. Shortcut connections enable the parameter gradients to flow more easily from the output layer to the earlier layers of the network.

Create the main branch of the network as a layer array. The addition layer sums multiple inputs element-wise. Specify the number of inputs for the addition layer to sum. To easily add connections later, specify names for the first ReLU layer and the addition layer.

layers = [
    imageInputLayer([28 28 1])
    
    convolution2dLayer(5,16,'Padding','same')
    batchNormalizationLayer
    reluLayer('Name','relu_1')
    
    convolution2dLayer(3,32,'Padding','same','Stride',2)
    batchNormalizationLayer
    reluLayer
    convolution2dLayer(3,32,'Padding','same')
    batchNormalizationLayer
    reluLayer
    
    additionLayer(2,'Name','add')
    
    averagePooling2dLayer(2,'Stride',2)
    fullyConnectedLayer(10)
    softmaxLayer
    classificationLayer];

Create a layer graph from the layer array. layerGraph connects all the layers in layers sequentially. Plot the layer graph.

lgraph = layerGraph(layers);
figure
plot(lgraph)

Create the 1-by-1 convolutional layer and add it to the layer graph. Specify the number of convolutional filters and the stride so that the activation size matches the activation size of the third ReLU layer. This arrangement enables the addition layer to add the outputs of the third ReLU layer and the 1-by-1 convolutional layer. To check that the layer is in the graph, plot the layer graph.

skipConv = convolution2dLayer(1,32,'Stride',2,'Name','skipConv');
lgraph = addLayers(lgraph,skipConv);
figure
plot(lgraph)

Create the shortcut connection from the 'relu_1' layer to the 'add' layer. Because you specified two as the number of inputs to the addition layer when you created it, the layer has two inputs named 'in1' and 'in2'. The third ReLU layer is already connected to the 'in1' input. Connect the 'relu_1' layer to the 'skipConv' layer and the 'skipConv' layer to the 'in2' input of the 'add' layer. The addition layer now sums the outputs of the third ReLU layer and the 'skipConv' layer. To check that the layers are connected correctly, plot the layer graph.

lgraph = connectLayers(lgraph,'relu_1','skipConv');
lgraph = connectLayers(lgraph,'skipConv','add/in2');
figure
plot(lgraph);

Load the training and validation data, which consists of 28-by-28 grayscale images of digits.

[XTrain,YTrain] = digitTrain4DArrayData;
[XValidation,YValidation] = digitTest4DArrayData;

Specify training options and train the network. trainNetwork validates the network using the validation data every ValidationFrequency iterations.

options = trainingOptions('sgdm', ...
    'MaxEpochs',8, ...
    'Shuffle','every-epoch', ...
    'ValidationData',{XValidation,YValidation}, ...
    'ValidationFrequency',30, ...
    'Verbose',false, ...
    'Plots','training-progress');
net = trainNetwork(XTrain,YTrain,lgraph,options);

Display the properties of the trained network. The network is a DAGNetwork object.

net

net = 
  DAGNetwork with properties:

         Layers: [16x1 nnet.cnn.layer.Layer]
    Connections: [16x2 table]
     InputNames: {'imageinput'}
    OutputNames: {'classoutput'}

Classify the validation images and calculate the accuracy. The network is very accurate.

YPredicted = classify(net,XValidation);
accuracy = mean(YPredicted == YValidation)

accuracy = 0.9932

Extended Capabilities

C/C++ Code Generation

Generate C and C++ code using MATLAB® Coder™.

Usage notes and limitations:

• Only the activations, classify, predict, predictAndUpdateState, classifyAndUpdateState, and resetState object functions are supported.

GPU Code Generation

Generate CUDA® code for NVIDIA® GPUs using GPU Coder™.

Usage notes and limitations:

• Only the activations, classify, predict, predictAndUpdateState, classifyAndUpdateState, and resetState object functions are supported.

Version History

Introduced in R2017b

R2024a: Not recommended

Starting in R2024a, DAGNetwork objects are not recommended, use dlnetwork objects instead.

There are no plans to remove support for DAGNetwork objects. However, dlnetwork objects have these advantages and are recommended instead:

• dlnetwork objects are a unified data type that supports network building, prediction, built-in training, visualization, compression, verification, and custom training loops.

• dlnetwork objects support a wider range of network architectures that you can create or import from external platforms.

• The trainnet function supports dlnetwork objects, which enables you to easily specify loss functions. You can select from built-in loss functions or specify a custom loss function.

• Training and prediction with dlnetwork objects is typically faster than LayerGraph and trainNetwork workflows.

To convert a trained DAGNetwork object to a dlnetwork object, use the dag2dlnetwork function.

This table shows some typical usages of DAGNetwork objects and how to update your code to use dlnetwork object functions instead.

Not Recommended	Recommended
Y = predict(net,X);	Y = minibatchpredict(net,X);
Y = classify(net,X);	scores = minibatchpredict(net,X); Y = scores2label(scores,classNames);
Y = activations(net,X,layerName);	Y = predict(net,X,Outputs=layerName);
[net,Y] = predictAndUpdateState(net,X);	[Y,state] = predict(net,X); net.State = state;
[net,Y] = classifyAndUpdateState(net,X);	[scores,state] = predict(net,X); Y = scores2label(scores,classNames); net.State = state;

To learn more about how to transition trainNetwork, SeriesNetwork, and DAGNetwork code to dlnetwork workflows, see Transition trainNetwork, SeriesNetwork, and DAGNetwork Code to dlnetwork Workflows.