An online retailer is offering a new line of running shoes. The retailer plans to send out an e-mail with a discount offer to some of its existing customers and wants to know if it can use data mining analysis to predict whether or not a customer might respond to its e-mail offer. The retailer prepares a data set of 170 existing customers who had received online promotions in the past, which include the following variables: Purchase (1 if purchase, 0 otherwise); Age (1 for 20 years and younger, 2 for 21 to 30 years, 3 for 31 to 40 years, 4 for 41 to 50 years, and 5 for 51 and older); Income (1 for $0 to $50K, 2 for $51K to $80K, 3 for $81K to $100K, 4 for $100K+); and PastPurchase (1 for no past purchase, 2 for 1 or 2 past purchases, 3 for 3 to 6 past purchases, 4 for 7 or more past purchases). A portion of the data set is shown in the accompanying table. 

a. Partition the data to develop a naïve Bayes classification model. Report the accuracy, sensitivity, specificity, and precision rates for the validation data set. 

b. Generate the decile-wise lift chart. What is the lift value of the leftmost bar? What does this value imply? 

c. Generate the ROC curve. What is the area under the ROC curve (or AUC value)? 

d. Can the naïve Bayes model be used to effectively classify the data? Explain your answer.

Transcribed Image Text:

Purchase Age 1 4 1 1 1 4 ⠀ 3 Income 3 1 ⠀ 4 PastPurchase 4 1 I 3