Understanding Feedforward Neural Network

In this assignment, you are required to build $3$ Feedforward Neural Networks$($FNNs$)$ to simulate

function

$y=7x_1^3-x_2*x_1^2*\frac{1-e^{e(x_1+1)x_2}}{1-e^{x_2+1}}$

in range $x1$:$+-10,2$:$+-10$

You need to have a different structure for each model you build.

Requirements

$1$ You are required to finish each part below following instructions.

$2$ You may work on native py files, but your work needs to include same documentation. You may

use pytorch or tensorflow keras.

$3$ You may work in a group of $3.$

For submission, submit a link to your github repo

Each student in the group must all submit.

In the cell below, generate training data for your model.

In the interval of $1$:$+-10,2$:$+-10,$ generate $1000$ data for $x1$ and $1000$ data for $2.$

All data need to be separated by the same distance.

That is$,$ you need to generate one value for $y$ for every $0\mathrm{.}02$ differences in $x1$ or $x2,$

and have a total of $1000000$ data in your training dataset.

Part $2$ build models