Homework in Cyper-physical and internet things. (see the attachment)

For this assignment, we will perform a reachability analysis for a feed-forward neural network using NNV (MATLAB Toolbox for Neural Network Verification). See the attached file for full information. Follow the exact instructions

Part A. Run the example.m script and capture the plot of input and output sets. Does the network satisfy the property of 1 < 5?

Part B. Modify example.m with new lower-bound and upper-bound vectors of input
set as [-2; -2] and upper-bound vector of [1; 1]. Run the modified script and capture the plot of
input and output sets. Does the network satisfy the property of 1 < 5?

Part C. Keep the same input bounds specified in Part B, now we modify example.m
with new weight matrices and bias vectors as follows
W1 = [5 2; 0.5 -2; 1 3]; b1 = [-1; 0; 0.5];
W2 = [-3 0.5 2; 0.5 2 -1]; b2 = [0.5; 1];
Run the modified script and capture the plot of input and output sets. Does the network satisfy the
property of 1 < 5?

Part D. Generate a random neural network with 2 inputs (with the same input
bounds specified in Part B), 1 hidden layer of 5 neurons and 2 outputs with arbitrary weight
matrices and bias vectors in a range of [-2 2]. Run the modified script and capture the plot of input
and output sets. Does your network satisfy the property of