EBay would like a method to predict which of its auction listings will be competitive (defined as having at bids from at least 2 different users) or not. The following variables are available for modeling:

Variable Description:
Category - One of 18 product categories, i. e. "Jewelry," "Automotive," etc.
Currency - One of 3 currencies: US dollar, Euro, or British pound
Seller Rating - The number of seller ratings. A continuous variable from 0 to 3,300
Duration - Duration in days. Takes the values 1, 3, 5, 7, or 10.
End Day - Day of the week on which the auction will end.
Close Price - The close price of the auction.
Open Price - The opening price of the auction.
Competitive? A dummy variable. 1 if the listing was competitive (2 or more bids were placed);
0 if the listing was not competitive (1 or fewer bids were placed).

a.) If eBay would like to use this algorithm to help sellers create competitive listings, are there any variables that should be excluded from the model? Explain. (Hint: Consider what information is available at the time the listing is posted.)

[Insert response.]

b.) List three possible classification algorithms that could be used. For each algorithm, include
i) A description of any data preparation that would be necessary prior to implementing the algorithm.
ii) An explanation of the advantages and disadvantages.

[Insert response.]