$2$ Using the backpropagation rule, train the opposite network for the data below in $2$ steps.

Suppose the activity function of nerves $($ is $ln$ this case, we have

$x=\left[\begin{array}{c}0\mathrm{.}5\\ 1\end{array}\right]y=\left[\begin{array}{c}1\\ -1\end{array}\right],x=\left[\begin{array}{c}0\mathrm{.}5\\ -0\mathrm{.}5\end{array}\right]y=\left[\begin{array}{c}1\\ 1\end{array}\right]$

Suppose all the initial weights of the network are equal to $0\mathrm{.}5$ get solution mataticly and explan it without code.