The Kohonen feature map module

The panel is divided into four sections: the Net Interface section with the buttons CONNECT , DISCONNECT and EXIT to allow the association between the network and the simulation module; the Simulation Scheduler with the buttons RUN , STEP, STOP , RND WEIGHT , EVAL , VAL TRSET , OPTION and some other tools to manage the simulations; the Training Set section with the buttons NEW , LOAD and IMAGE to manage the training set generation and the training set files; the forth section will show output data of the simulations and context dependent messages. The application guides the use of the panel changing the sensitivity of the panel tools according to the feasible operations at a given phase of the simulation.

The Net Interface section

CONNECT: It establishes the connection between the model panel and the current net; the net is tested to see if the structure is compatible with the kohonen model architecture.
DISCONNECT: It erase the connection to the net, deallocating from memory the simulation data.
EXIT: It closes the Hebbian simulation panel.

The Simulation Scheduler section

RUN

It starts/restarts the simulation on the current net, with current parameters and options settings. While simulation is running, the STOP and STEP buttons are sensitive.

STEP

It is used to perform simulations step by step: when used the simulation runs for a number of steps indacated by the refresh parameter.

STOP

Allows thr user to stop the running simulation; this feature is useful because the simulation time could be quite large, expecially with big nets (hundreds of neurons), and the user could think to change the simulation parameters and restart.

RND WEIGHT

This button allows to set random values to the arcs weight: this is useful when the user would restart a new simulation.

EVAL

When the learning is terminated the user sets the input neurons whit some value. The EVAL button evaluates the output values using the arcs weight learned before.

OPTIONS

Clicking on this button will open an option panel with some Motif toggle buttons, which allows to set some simulation flags to modify both model behaviour and output features.

Conscience :: it enables a method to avoid dead units called conscience when the simulation runs.
Lambda Form :: the user can choose the form of the neighborood function lambda .
Report File :: selecting yes an output text file is written, while the simulation is running. The output file can be found in directory <INNEHOME>/NOTES and is named Kohonenreport.txt (a successive output will rewrite the last one); the output file reports different statistics on data taken after each iteration of the learning algorithm.
Eta Rule : Costant, Linear, Armonic, Exp: there are four eta decay rules: with Costant rule, the eta value does not change in the simulation; with the other rules, eta will decay according to the selected formula.
Sigma Rule : Costant, Linear, Armonic, Exp: like eta rule, there are four sigma decay rules.
Norm : 1, p, infinite: this option sets the norm type. The norm can be 1, p or infinite.

The Simulation Parameters

Initial Eta Value: It is the value used for eta by the first simulation step. While the simulation runs, the eta value decays according to the Eta Rule .
Initial Sigma Value: It is the value used for sigma by the first simulation step. While the simulation runs, the sigma value decays according to the Sigma Rule .
Learning Cycles: It represents the number of steps performed in the simulation.
Refresh Frequence: It is the number of steps between two video refreshes.

The Training Set section

NEW: If the net has two input neurons, it is possible to build a bidimensional polygon where uniform distribution of points is generated. To fix a vertex click the left button The right button ends the polygon costruction. A polyline can be made by clicking the central button.
LOAD: Loads from a training set file a sequence of example; the input neurons determines the expected format of the file; if this format is not satisfied, the load fails.

Back to index