Neural Network Design and Analysis:

Perceptron $($weighted sum & threshold$)$ model

The feed$-$forward network shown below uses bipolar neurons, and maps the entire plane $x_1,x_2,$ into a bipolar value. Find the segment of the plane for which $y_4=1.$ NOTE: the neurons in the first computational layer have the neuronal $($activation$)$ function shown below. The neuron in the output layer is an MP neuron.

$f(y_{in})=\{\begin{array}{ccc}+1,& if& y_{in}0\\ -1,& if{y}_{-}in& <mn>0</mn>\end{array}$

Design a feed$-$forward network composed of bipolar neurons that will distinguish between the data points in Class $1($O$\text{'}$s$)$ and Class $2($X$\text{'}$s$).$ HINT: it is easiest to use the activation function:

$f(y_{in})=\{\begin{array}{ccc}+1,& if& y_{in}0\\ -1,& if{y}_{-}in& <mn>0</mn>\end{array}$