In this assignment, we will use the auctions data you have seen before. Note that there are $3$ separate datasets below for training, testing and new observations $($includes auctions for which we don't know competitiveness$).$ Follow the steps:

$(5$ points$)$ Use the training dataset to train a Naive Bayes Classifier to predict Competitive based on all variables.

Show model output for all variables except category

What are the important explanatory variables? Use variable importance plot

$(5$ points$)$ Use the test dataset to check for model performance.

Show the confusion matrix table for test data

Discuss all the performance metrics we talked about so far, such as accuracy, precision, F$1$ score, AUC etc.

$(5$ points$)$ Predict 'competitive' for auctions in the new observations dataset. Remember you can export your predictions using the blue icon on the top right in radiant.

How many auctions are competitive vs$.$ not competitive if cutoff is $0\mathrm{.}5?$

What about if cutoff is $0\mathrm{.}7?$ Does this make sense?