Tools

To understand and implement a Multi-Layer Perceptron (MLP) with focus on architecture, activation functions, and training

  1. Click on CREATE tab.

  2. Enter the target value for the XOR gate in the table, then click the SUBMIT button.

  3. Navigate to the COMPUTE tab in the header and click on it.

  4. Initiate the process by clicking the GENERATE WEIGHTS button to initialize weight values.

  5. Proceed to select values for the input nodes, specifically X1 and X2. Confirm your selections by clicking the SUBMIT button.

  6. Identify the Zᵢ cell corresponding to the current row and input the appropriate Zᵢ value.

  7. Click the ACTIVATION FUNCTION button.

  8. Confirm result accuracy by clicking the CHECK button to compare target (X₁.X₂') and output values (f(Zᵢ)).

  9. If X₁.X₂' does not match f(Zᵢ), follow these steps. Else, jump to step 10:

    1. Click the UPDATE button to adjust weights.
    2. Enter the learning rate value.
    3. Provide necessary values in the input boxes for W11 and click SUBMIT button.
    4. Repeat the above step (9.3) for W12.

  10. Repeat steps 5 through 9 for all four combinations of X1 and X2 .

  11. Proceed by clicking the NEXT button to execute epochs until the predicted output is obtained for each input dataset.

  12. Lastly, proceed to the ANALYZE tab and select it. Generate the graph by clicking the PLOT button.