com.practicalstudies.neural
Class WeightedConnection

java.lang.Object
  |
  +--com.practicalstudies.neural.WeightedConnection

public class WeightedConnection

extends java.lang.Object

Represents a connection between Nodes.

A WeightedConnection has three attributes:

• an input Node;
• an output Node; and,
• a connection weight which will be used to modify the output of the input Node before it is delivered to the output node.

Given that a WeightedConnection has an explicit input-output direction, the combination of a group of WeightedConnections and Nodes forms a directed graph.

Author:

Michael Wax

Constructor Summary

WeightedConnection()
Constructor.

WeightedConnection(Node inputNode, Node outputNode)
Constructor which allows assignment of input and output Nodes to the connection.

Method Summary

void	adjustWeight() Called to adjust the connection weight of this connection if some learning algorithm is in place.
Node	getInputNode()
double	getLastWeightChange() Returns the last change in weight of this function during network learning.
Node	getOutputNode()
double	getWeight() Returns the weight of this connection.
WeightAdjustmentFunction	getWeightAdjustmentFunction()
void	randomizeWeight() Assign a random value to the connection weight of this connection.
void	setInputNode(Node inputNode)
void	setLastWeightChange(double lastWeightChange) Specify the last change in weight of this function.
void	setOutputNode(Node outputNode)
void	setWeight(double weight) Specify the weight of this connection.
void	setWeightAdjustmentFunction(WeightAdjustmentFunction weightAdjustmentFunction)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

WeightedConnection

public WeightedConnection()

Constructor.

Calls randomizeWeight() to assign a random connection weight to the connection.

WeightedConnection

public WeightedConnection(Node inputNode,
                          Node outputNode)

Constructor which allows assignment of input and output Nodes to the connection.

Calls randomizeWeight() to assign a random connection weight to the connection.

Parameters:

inputNode - the Node which will provide input to this connection

outputNode - the Node which will receive input through this connection

Method Detail

randomizeWeight

public void randomizeWeight()

Assign a random value to the connection weight of this connection.

adjustWeight

public void adjustWeight()

Called to adjust the connection weight of this connection if some learning algorithm is in place.

The learning algorithm will be encapsulated in a WeightAdjustmentFunction as specified by setWeightAdjustmentFunction.

As an alternative to using this method, the weight may be set explicitly using the setWeight method.

getWeight

public final double getWeight()

Returns the weight of this connection.

setWeight

public final void setWeight(double weight)

Specify the weight of this connection.

getLastWeightChange

public final double getLastWeightChange()

Returns the last change in weight of this function during network learning.

setLastWeightChange

public final void setLastWeightChange(double lastWeightChange)

Specify the last change in weight of this function.

getInputNode

public Node getInputNode()

setInputNode

public void setInputNode(Node inputNode)

Parameters:

inputNode - the Node which will provide input to this connection

getOutputNode

public Node getOutputNode()

setOutputNode

public void setOutputNode(Node outputNode)

Parameters:

outputNode - the Node which will receive input through this connection

getWeightAdjustmentFunction

public WeightAdjustmentFunction getWeightAdjustmentFunction()

setWeightAdjustmentFunction

public void setWeightAdjustmentFunction(WeightAdjustmentFunction weightAdjustmentFunction)