Directed acyclic graph (DAG) network for deep learning
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.
There are several ways to create aDAGNetwork
object:
Load a pretrained network such assqueezenet
,googlenet
,resnet50
,resnet101
, orinceptionv3
. For an example, seeLoad SqueezeNet Network. For more information about pretrained networks, seePretrained Deep Neural Networks.
Train or fine-tune a network usingtrainNetwork
. For an example, seeTrain Deep Learning Network to Classify New Images.
Import a pretrained network from TensorFlow™-Keras, TensorFlow 2, Caffe, or the ONNX™ (Open Neural Network Exchange) model format.
For a Keras model, useimportKerasNetwork
. For an example, seeImport and Plot Keras Network.
For a TensorFlow model in the saved model format, useimportTensorFlowNetwork
. For an example, seeImport TensorFlow Network as DAGNetwork to Classify Image.
For a Caffe model, useimportCaffeNetwork
. For an example, seeImport Caffe Network.
For an ONNX model, useimportONNXNetwork
. For an example, seeImport ONNX Network as DAGNetwork.
Assemble a deep learning network from pretrained layers using theassembleNetwork
function.
Note
To learn about other pretrained networks, seePretrained Deep Neural Networks.
activations |
Compute deep learning network layer activations |
classify |
Classify data using a trained deep learning neural network |
predict |
Predict responses using a trained deep learning neural network |
plot |
Plot neural network layer graph |
trainNetwork
|trainingOptions
|importKerasNetwork
|layerGraph
|classify
|predict
|plot
|googlenet
|resnet18
|resnet50
|resnet101
|inceptionv3
|inceptionresnetv2
|squeezenet
|SeriesNetwork
|analyzeNetwork
|assembleNetwork