Neural Networks

Consider the following neural network with a step function (of threshold =0) for the activation function.

Note that x₁ and x₂ are input variables, and y is the output (class).

Step function with threshold = 0:

The training data consists of 5 data points as illustrated in the following figure. Red circle represents the positive class (y=1) and blue circle the negative class (y=0).

(1)Write two linear equations for the neural network decision boundary. Draw on the figure the neural network decision boundary.

(2)What is the training error? Show your computations.