Class Model object for the training stage of neural network.
More...
◆ TrainingModel() [1/2]
Constructs model for the training stage of the neural network
- Parameters
-
context | Context to manage the model |
◆ TrainingModel() [2/2]
Constructs model for the training stage of the neural network by copying parameters of another model
- Parameters
-
context | Context to manage the model |
other | A model to be used as the source to initialize the parameters of the model |
◆ getBackwardLayer()
Returns the backward stage of a layer with certain index in the network
- Parameters
-
index | Index of the layer in the network |
- Returns
- Backward stage of a layer with certain index in the network
◆ getBackwardLayers()
Returns list of backward layers
- Returns
- List of backward layers
◆ getForwardLayer()
Returns the forward stage of a layer with certain index in the network
- Parameters
-
index | Index of the layer in the network |
- Returns
- Forward stage of a layer with certain index in the network
◆ getForwardLayers()
Returns list of forward layers
- Returns
- List of forward layers
◆ getNextLayers()
Returns list of connections between layers
- Returns
- List of next layers for each layer with corresponding index
◆ getPredictionModel()
Returns list of forward layers and their parameters organised in the PredictionModel
- Parameters
-
cls | Data type to use in intermediate computations for the neural network, Double.class or Float.class |
- Returns
- List of forward layers and their parameters organised in the PredictionModel
◆ getWeightsAndBiases()
Returns the weights and biases of the forward layer of neural network as numeric table
- Returns
- The weights and biases numeric table
◆ initialize()
void initialize |
( |
Class<? extends Number > |
cls, |
|
|
long [] |
sampleSize, |
|
|
TrainingTopology |
topology |
|
) |
| |
Initializes neural network
- Parameters
-
cls | Data type to use in intermediate computations for the neural network, Double.class or Float.class |
sampleSize | Dimensionality of the training data |
topology | Collection of layer descriptors of every inserted layer |
◆ setWeightsAndBiases()
Sets the weights and biases of the forward layer of neural network as numeric table
- Parameters
-
weightsAndBiases | The weights and biases numeric table |
The documentation for this class was generated from the following file: