The enclosed xlsx file contains data on over 3000 customers of a cell phone provider. Columns B through N provide information about the account and usage of each customer, and column O indicates whether the customer churned during the given period of time.
Use the Discriminant Analysis (see the Hint below), the enclosed data, and the Excel Solver to develop a classification model for predicting whether a customer will churn.
Hint: The Discriminant Analysis assigns a weight vector to the data (features), and then computes the weighted combination of the data for each customer. The customer’s weighted combination is called the customer’s discriminant score. The method then classifies a customer into group 1 or group 2 depending on which side of a cutoff score the customer’s discriminant score falls. computes the weighted combination of the data for each customer. The problem is to find the optimal weight vector and cutoff score that maximize the percentage of correct classifications in the training data.