01 / The brief
The question
Build a classification model that helps identify customers at risk of leaving the bank.
This GitHub project focuses on predicting whether a bank customer is likely to churn. It is relevant for the portfolio because it shows practical machine learning skills for a real business problem: customer retention.
Results & observations
- Created a churn prediction workflow that supports customer retention analysis.
- The project demonstrates classification, feature preparation, and model evaluation skills.
Challenges
- Classifying customers correctly
- Preparing mixed customer features
- Choosing useful evaluation metrics
- Explaining model results in business terms
Approach
- Prepared customer churn data for machine learning.
- Handled numerical and categorical features.
- Trained classification models.
- Evaluated model performance with relevant metrics.
Implementation details
Data Preparation
Prepared the bank churn dataset for machine learning.
Feature Engineering
Processed customer features for model training.
Model Training
Trained classification models to predict churn.
Model Evaluation
Evaluated the model using classification metrics.
Business Interpretation
Connected churn predictions to customer retention decisions.
Explore the report
Select a screenshot for a larger view.