DAGNetwork

Encapsulates the details of a network with a directed acyclic graph structure.

This class does not provide facilities to actually train the network---that can be accomplished with the dopt.online package.

class DAGNetwork {

this(Operation[] inputs, Layer[] outputs);

Operation[] inputs();

Operation[] outputs();

Operation[] trainOutputs();

Operation paramLoss();

Projection[Operation] paramProj();

Operation[] params();

}

Constructors

this this(Operation[] inputs, Layer[] outputs): Construct a DAGNetwork with the given inputs and outputs.

Members

Functions

inputs Operation[] inputs(): The inputs provided when the DAGNetwork was constructed.
outputs Operation[] outputs(): The Operation objects produced by the output layers provided during construction.
paramLoss Operation paramLoss(): The sum of all the parameter loss terms.
paramProj Projection[Operation] paramProj(): An associative array of projection operations that should be applied to parameters during optimisation.
params Operation[] params(): An array of all the Operation nodes in the graph representing network parameters.
trainOutputs Operation[] trainOutputs(): Separate Operation objects produced by the output layers provided during constructions.

Meta

Source

See Implementation

dopt nnet networks modules