Assignment 9
Jenna Sempkowski & Victoria Apostolico
2024-04-11
Overview of the Data
- There were 10127 observations in the CreditCardData dataset.
- We split our data 70/30.
- We will be predicting the customers that leave and the customers that stay.
- There are 19 predictors.
Variable Importance
Accurary of the model
##
##
## Cell Contents
## |-------------------------|
## | N |
## | Chi-square contribution |
## | N / Row Total |
## | N / Col Total |
## | N / Table Total |
## |-------------------------|
##
##
## Total Observations in Table: 3072
##
##
## | val$Attrition_Flag
## val$predicted_rf | Attrited Customer | Existing Customer | Row Total |
## ------------------|-------------------|-------------------|-------------------|
## Attrited Customer | 406 | 25 | 431 |
## | 1598.842 | 312.303 | |
## | 0.942 | 0.058 | 0.140 |
## | 0.809 | 0.010 | |
## | 0.132 | 0.008 | |
## ------------------|-------------------|-------------------|-------------------|
## Existing Customer | 96 | 2545 | 2641 |
## | 260.924 | 50.967 | |
## | 0.036 | 0.964 | 0.860 |
## | 0.191 | 0.990 | |
## | 0.031 | 0.828 | |
## ------------------|-------------------|-------------------|-------------------|
## Column Total | 502 | 2570 | 3072 |
## | 0.163 | 0.837 | |
## ------------------|-------------------|-------------------|-------------------|
##
## Financial Impact of the Model
We predicted 508 were going to leave.
The cost of my program is 508 x 100 = $50,800
We will save 30% of 406 = 121.8, which has a value of 2300 x 121.8 = $280,140
The net of this program is $229,340 from 2990.
Therefore, it is $75.22 per customer.