Assignment 9
Libby Camardella
2024-04-11
Data Preparation
- Split the data (10127), into training and validation datasets (70%/30%, 7234/2893).
Variable Importance
Model Accuracy
Using the validation data, it is observed that the model is able to:
Predict 421 customers were going to leave and 2472 customers were going to stay.
488 customers actually left and 2405 customers actually stayed.
The financial impact of the model, given the following assumptions:
The customer retention program consists of forgiving the anual fee of $100 (on average) = 421($100) = $42,100
The value of a customer is $2,300 per year (on average) = 118.8($2300) = $273,240
The customer retention programs have a 30% success rate 396(.30) = 118.8
Out of 2893 customers, this model will generate $231,140, or roughly $80 per customer.
##
##
## Cell Contents
## |-------------------------|
## | N |
## | Chi-square contribution |
## | N / Row Total |
## | N / Col Total |
## | N / Table Total |
## |-------------------------|
##
##
## Total Observations in Table: 3056
##
##
## | val$predicted_Attrition_Flag
## val$Attrition_Flag | Attrited Customer | Existing Customer | Row Total |
## -------------------|-------------------|-------------------|-------------------|
## Attrited Customer | 391 | 100 | 491 |
## | 1545.822 | 246.980 | |
## | 0.796 | 0.204 | 0.161 |
## | 0.929 | 0.038 | |
## | 0.128 | 0.033 | |
## -------------------|-------------------|-------------------|-------------------|
## Existing Customer | 30 | 2535 | 2565 |
## | 295.906 | 47.278 | |
## | 0.012 | 0.988 | 0.839 |
## | 0.071 | 0.962 | |
## | 0.010 | 0.830 | |
## -------------------|-------------------|-------------------|-------------------|
## Column Total | 421 | 2635 | 3056 |
## | 0.138 | 0.862 | |
## -------------------|-------------------|-------------------|-------------------|
##
##