Make your own free website on Tripod.com

Interface Description

The yellow bar tells you what to do.

The 4 buttons:

1. User Training: Select and train the neuron by yourself.
2. Slow System Training: Select and slow you the steps of default training.
3. Fast Traininig: Select and training by default without showing the steps.
4. Reset: Can return the program to the original status.

There is a online help link which brings you to this website :).

Input Training Sets: (Activiate only when user training or relearn is press)

1. The 6 boxs allow you to type the vector.
2. The scroll bars allow you to pull up and down for input.
3. The 2 orange buttons is for the target output.

User Input: (Activiate only when training is finished)

1. The six scroll bars allow you to input the vector for classify. The alternative is typing by user in the neuron diagram.
2. The Classify button is used to call the neuron to classy the curve.
3. If the classify is incorrect, you can press Relearn button to train again.

Weight Adjust: (Never activiate)

1. Target Output: Shows you the target output
2. Actual Output: Shows you the actual output from the classification of the neuron.
3. The black box shows if the target output is equal to the actual output. Y for yes and N for no.
4. The table shows the adjustment of weights. w1 at the top and θ at the buttom.
5. The formula is w + Δw = new w.

Neuron Diagram: The diaram shows the training and classification of the curve.

1. x1 to x6 is the input vector and it will activate when training is finished.
2. The new weight will be shown on w1 to w6 and θ boxes.
3. The result of v and y will be shown.
4. The black box will show the result of classification, '∩' or 'U' will be displayed or '-' for the horizontal lines.