RFI II activation fiunctions

$N_1-u.J$

$b_2=-0\mathrm{.}2$

a$)$ For given inputs, weights and biases calculate the feed forward parameters $(s_1,s_2,u_1,u_2$ and hat$(y))$ of the Neural Network. $($Please give calculation details$)$

b$)$ Considering Squared Error Loss, $L_i=\frac{1}{2}(y_i-$hat$(y{)}_i{)}^2,$ obtain the upstream gradient to be used first in the Backpropagation algorithm to calculate the derivatives.$($Please give calculation details$)$

c$)$ Assuming the target output for input pairs $)=1,x_2=(1$ is $y=1$ and using the Squared Error Loss in $($b$),$ apply the Backpropagation algorithm to calculate the derivatives of the Loss function with respect to weights and biases. $($Please give calculation details$)$

d$)$ Using the derivatives calculated in $($c$)$ in Gradient Descent Algorithm, update the weights and biases of the Neural Network $($Assume learning parameter is $=0\mathrm{.}1).($Please give calculation details$)$