Create sample data using two classes that are separable linearly and implement the perceptron to determine an appropriate decision function.

Repeat this process and include data samples that will make the two classes not linearly separable and see if the Perceptron still works albeit with a lesser classification accuracy.

Implement the generalized decision function formulation $($GDF$)$ for the rth order and with n Illustrate that your program works for the cases:

$($a$)$ Second degree with $2$ dimensions

$($b$)$ Third degree with $2$ dimensions

$($c$)$ Second degree with $3$ dimensions

$($d$)$ Third degree with $3$ dimensions

Perhaps initially you set all the weights to $1$ and observe your results, then as you vary the different weights singularly$,$ look at the change experienced in the various GDFs$.$

Use Jupyter notebook python