infile = "WAFnUseCTelcoCustomerChurn.csv"

Consolidations : Customer Churns in Telecom Industry

Customer Churns in Telecom Industry : Decision Tree Analysis

Decision Tree Model : Output Summary 

indata_model_Telecom
## 
## Model formula:
## Churn ~ gender + PhoneService + MultipleLines + StreamingMovies + 
##     Contract + PaymentMethod
## 
## Fitted party:
## [1] root
## |   [2] Contract in Month-to-month
## |   |   [3] StreamingMovies in No, Yes
## |   |   |   [4] PaymentMethod in Bank transfer , Credit card , Mailed check: No (n = 949, err = 36.9%)
## |   |   |   [5] PaymentMethod in Electronic check: Yes (n = 1073, err = 45.2%)
## |   |   [6] StreamingMovies in No internet service: No (n = 312, err = 16.7%)
## |   [7] Contract in One year, Two year
## |   |   [8] Contract in One year
## |   |   |   [9] StreamingMovies in No, No internet service: No (n = 493, err = 4.5%)
## |   |   |   [10] StreamingMovies in Yes: No (n = 386, err = 19.2%)
## |   |   [11] Contract in Two year
## |   |   |   [12] PaymentMethod in Bank transfer , Credit card , Mailed check: No (n = 913, err = 1.4%)
## |   |   |   [13] PaymentMethod in Electronic check: No (n = 99, err = 7.1%)
## 
## Number of inner nodes:    6
## Number of terminal nodes: 7

Decision Tree Model : Confusion Matrix 

indata_pred_Telecom = predict(indata_model_Telecom, indata_test)
Telecom_conft = table("prediction" = indata_pred_Telecom, "actual" = indata_test$Churn)
Confusion Matrix: Test Data: Customer Churns
Predicted
Actual Class
No Yes
No 1740 370
Yes 315 393