Question $2($Support Vector Machines$)$:

In google collab $-$ python

a$)$Generate a scatter plot of the data in $$customers$\_$data.csv$.$ Color code the data points to visualize the datapoints belonging to each class.

b$)$Preprocess the data and build an SVM classifier with very large C value. What type of margin classifier is this?

c$)$Plot the decision boundary of the resulting SVM model and show the margins and the resulting support vectors. How many support vectors are there?

d$)$Evaluate the classifier on the $$customers$\_$data$\_$test.csv$.$ Report the value$($s$)$ of the appropriate evaluation metric$($s$).$

e$)$Use the $$customers$\_$data$\_$full.csv$$ dataset to build an SVM classifier. Set the value of C to be a very large value.

f$)$Discuss any issues you run into when building an SVM classifier using $$customers$\_$data$\_$full.csv$$ and a large value for C$.$ Explain what could be causing those issues.

g$)$Build an SVM classifier that is suitable for the data in $$customers$\_$data$\_$full.csv$.$ What type of margin classifier is this?

h$)$Choose the value of C that results in the best performance on $$customers$\_$data$\_$test.csv$.$

i$)$Plot the decision boundary of the resulting SVM model and show the margins and the resulting support vectors. How many support vectors are there?