Click or drag to resize

BackpropagationSettings Structure

The configuration of the algorithm.

Namespace:  RavingBots.MagicGestures.AI.Neural.Classic
Assembly:  Assembly-CSharp (in Assembly-CSharp.dll) Version: 0.0.0.0
Syntax
C#
Copy
public struct Settings

The BackpropagationSettings type exposes the following members.

Properties
  NameDescription
Public propertyStatic memberDefault
The default settings.
Public propertyIsValid
if the chosen settings are correct.
Top
Fields
  NameDescription
Public fieldLearningRate
The learning rate is used by the backpropagation training algorithm. The parameter determines the speed of the training.
Public fieldMaxIterations
The training is continued for the specified number of iterations unless the target error is reached earlier.
Public fieldMomentum
The momentum parameter is used by the backpropagation training algorithm. The parameter accelerates the training process.
Public fieldRandomWeightRange
The range from which the initial random weights are selected.
Public fieldTargetError
The training process is stopped when the training error reaches the specified value.
Top
Remarks
The algorithm runs for a set number of iterations (MaxIterations) or until the error value drops below a set target (TargetError). The speed and the accuracy of the training are tuned by two parameters: the learning rate (LearningRate) which determines whether the network should learn faster or more accurately, and the momentum (Momentum) which determines how much a weight update should additionally depend on a previous change.
See Also