1. Click on the CREATE tab.
2. Select the number of nodes for all three layers and click the SUBMIT button.
3. Click on NEXT button to proceed further.
4. Choose an input value for each node in the input layer and an output value for each node in the output layer, then click the SUBMIT button.
5. Click the NEXT button to initialize the weights.
6. Choose the source (Xᵢ) and destination (Hᵢ), enter the weight value, then click the SUBMIT button.
7. Repeat the previous step (6) for every combination of source (Xᵢ) and destination (Hᵢ).
8. Now, select the source and destination to connect the hidden layer to the output layer. Enter the weight value and click the SUBMIT button.
9. Repeat step 8 for every combination of source (Hᵢ) and destination (Yᵢ).
10. Click the NEXT button to compute the forward propagation.
11. Enter the bias value (random) and click the SUBMIT button.
12. For each node in the hidden layer, calculate the summation by entering the appropriate input and weight values, then click the SUBMIT button.
13. Repeat the summation process for each node in the output layer, using the same method.
14. Finally, click the ERROR button to compute the Mean Squared Error (MSE).

Formulae:

1. To compute y_in for an activation function :
   
2. Sigmoid activation function :
   
3. Mean Squared Error (MSE) :
   

Note: In our simulation, we used the same bias for every node; however, in practice, biases are often adjusted individually to allow for more flexibility in learning complex patterns.