[This is preliminary documentation and is subject to change.]

This is the base parameters class for neural network classes that derive from NeuralNetwork

Namespace: Clarion.Framework.Templates
Assembly: CLARIONLibrary (in CLARIONLibrary.dll) Version: 6.1.0.7 (6.1.0.7)

Syntax

C#
public class NeuralNetworkParameters : ImplicitComponent..::..ImplicitComponentParameters

Remarks

This class contains all of the parameters that may be used by the NeuralNetwork class. All classes that extend NeuralNetwork should also contain their own parameters class that extends this class and adds any parameters that are used by those classes. This is how all of the "built-in" neural networks within the CLARION Library have been implemented. It is HIGHLY recommended that users defining their own (custom) neural network follow this convention as well. Details on how to implement a parameters class for a custom neural network can be found in the Advanced Tutorial under the Tutorials section of the CLARION Library package.

Note

The parameters classes can be found throughout the CLARION Library whenever a class contains parameters that may be set by a user. This concept was implemented to allow for some "special capabilities" and also to clean-up/streamline parameter usage throughout the system. The parameters classes enable parameters to be adjusted both locally and globally. In general, global parameter changes are performed statically through the GlobalParameters property and local parameter changes are performed on single instances through the Parameters property.

Examples demonstrating how parameter adjustments may be performed using both methods can be found in the examples section of ClarionComponent..::..ClarionComponentParameters.
Note
Global parameter settings are ONLY applicable prior to the initialization of an instance of a neural network. During the initialization process, the global parameter settings are used to set the "default" values of the local parameters for a neural network. Once initialized, the neural network will only use the local parameters. Therefore, global parameter changes SHOULD ALWAYS be performed BEFORE instances of a neural network are initialized.

Copyright 2011. Nicholas Wilson

Examples

Inheritance Hierarchy

See Also