Given the Artificial Neural Network model displayed as above, Let $f_1(e)=0\mathrm{.}2,f_2(e)=-0\mathrm{.}5$ and $f_3(e)=0\mathrm{.}3$ be the output values of internal teurons in the first hidden layer, and let $w_{14}=0\mathrm{.}3,w_{15}=0\mathrm{.}3,w_{24}=0\mathrm{.}5,w_{25}=0\mathrm{.}6,w_{34}=0\mathrm{.}6$ and $w_{35}=0\mathrm{.}3$ be weights of edges connecting these hidden neurons. The functions $f_4(e)$ and $f_s(e)$ are defined as $f_4(e)=f5(e)=$ReLU$(e)=$ max$\{0,e\}.$

For the Feedforward Neural Network, your task is to calculate the output of $f_4(e)$ and $f_5(e)$

Considering the back propagation algorithm. If the errors of $f_4(e)$ and $f_5(e)$ are $z_4=0\mathrm{.}04$ and $z_5=0\mathrm{.}2,$ respectively, what is the error back propagated to internal neurons $f_1(e)$ and $f_2(e)$ with the back propagation algorithm.

You need to show your calculation steps.