Creating the map |
Mapping is done before you can label your network.
Press the Map button and the map is drawn to the dimensions that you specified when creating the network.
This file can be saved and then viewed.
Saving the map |
After creating the new map you should save it.
To do this select Save As Map from the Operations menu.
Here you are able to choose what directory you want to save your network in, and specify the name of the file.
The convention for the file extension of map files is *.map.
The only restriction is that the path cannot contain a ).
You can now view the map file.
Viewing the map file |
To view the map file select View Map from the Operations menu.
In the map file you should see the coordinates of the node on the map and then a label followed by a number. This number represents the number of times that node was activated by the class identified by the label.
For example:
[FormatVersion = 1.0]
[DataSet]
[Rows = 20]
[Inputs = 2]
[ExpectedOutputs = Labels]
[GeneratedOutputs = 0]
[Error = 0]
[Data]
0 0
[IrisSetosa=7.000000000 ] /*Node (0,0) was activated by
IrisSetosa 7 times*/
0 1 [ ] /*Node (0,1) was
not activated by any of the classes*/
0 2
[IrisVersicolor=5.000000000 IrisVirginica ] /*Node (0,2)
was activated by IrisVersicolor 5 times and IrisVirginica once*/
/*... more data would be here*/
[~Data]
[~DataSet]
Plotting the results |
After recalling and labeling the network you can select Plot from the Operations menu. This will plot the results of training the network.
Unfortunately the Plot function is not working yet so you can only view the file through a text editor.