This project includes a machine learning model developed to predict the likelihood of customers churning in a bank.
- Total Features : 10
- Total Row : 10000
- CSV File Size : 837.42 kB
| Feature | Description |
|---|---|
| RowNumber | Sequential record number with no impact on the analysis or outcome. |
| CustomerId | Unique identifier assigned to each customer, devoid of influence on customer attrition. |
| Surname | Family name of the customer, considered irrelevant in assessing the likelihood of them leaving the bank. |
| CreditScore | Numeric value reflecting the customer's creditworthiness, influencing the probability of churn; higher scores indicate lower likelihood. |
| Geography | The geographical location of the customer, contributing to the analysis as it may affect their decision to leave the bank. |
| Gender | Exploration of whether gender plays a role in customer churn. |
| Age | A relevant factor, as older customers tend to exhibit greater loyalty and are less prone to leaving the bank compared to younger counterparts. |
| Tenure | The duration in years that the customer has been a client of the bank, generally associated with increased loyalty. |
| Balance | A crucial indicator of customer churn; higher account balances correlate with lower likelihood of departure. |
| NumOfProducts | The count of products purchased by the customer from the bank, influencing their overall engagement. |
| IsActiveMember | Indicator of an active customer, with active members less likely to leave the bank. |
| EstimatedSalary | Similar to balance, lower salaries are associated with higher probabilities of customer churn. |
| Exited | Binary indicator representing whether the customer left the bank (1) or not (0). |
| Complain | Binary indicator denoting whether the customer has lodged a complaint (1) or not (0). |
| Satisfaction Score | Score provided by the customer reflecting their satisfaction with the bank's complaint resolution process. |
| Card Type | The type of card held by the customer, potentially impacting their engagement and loyalty. |
| Points Earned | The points earned by the customer through credit card usage, contributing to the overall customer profile. |
Model : LinearClassifier (Tensorflow)
Accuracy : %99