Consider a neural network with a single hidden layer with two neurons, two inputs x$1$ and x$2,$ and activation function f$($x$)=$ e$^$x$/(1+$e$^$x$)$ for all the neurons. This network has to be trained for binary classification by using the training data set $($xi$1,$ xi$2,$ Yi$),$ i $=1,2,...,$ n$,$ where Y$1,$ Y$2,...,$ Yn are the class labels in the training data. What is the expression for the task of minimizing the squared error loss function through the backpropagation algorithm? Give the detailed expression and explain all the steps.