Sample Run

Here is a sample run of what happens under the following settings:

Data

• Dataset: Circle
• Ratio of training to test data: 50%
• Noise: 0
• Batch Size: 10

Inputs

• Features: X1 and X2

Hidden Layers

• 2 Hidden Layers
• First Layer: 4 neurons
• Second Layer: 2 neurons

Parameters

• Learning Rate: 0.3
• Activation: Tanh
• Regularization: None
• Regularization Rate: 0
• Problem Type: Classification

The neural network ran for 215 iterations, and we get the following window:

We see that the positive data is clustered in a region of blue while the negative data is clustered in a region of orange. This means the neural network successfully classified most (if not all) of the data points.