Problem $2.$ Implement back$-$propagation algorithm for following functions $($do not use matlab toolbox for this problem$)$

$f(x)=e^{-x}$ where $1x10$

$f(x,y)=sin(x)cos(y)$ where $n(-2,2)$ and yin$(-2,2)$

For each function, do the following:

Set up three sets of data: training, validation and testing $($already provided in the code$).$

Use the training and validation data to compute the synaptic weights of the network, assumed to have a single hidden layer. The training data should be used to update the weights and validation data to calculate the MSE.