Credit Card Customer Segmentation

This project aims to perform customer segmentation analysis on credit card users to identify distinct customer groups based on their behavior and characteristics. The analysis uses unsupervised machine learning techniques, specifically K-means clustering, to group customers into meaningful segments.

Project Overview

The project analyzes various customer attributes such as spending behavior, credit limits, payment patterns, and other relevant features to identify natural groupings of customers. This segmentation can help in:

• Developing targeted marketing strategies
• Improving customer service
• Risk management
• Product development
• Understanding customer behavior patterns

Setup

1. This project uses Poetry for dependency management. Ensure you have Poetry installed:

curl -sSL https://install.python-poetry.org | python3 -

2. Create a new virtual environment and install dependencies:

poetry env use python3
poetry install

3. Run Python scripts or commands within the Poetry environment:

poetry run python your_script.py

# Or run Jupyter notebook
poetry run jupyter notebook

Note: Instead of activating the virtual environment directly, we use poetry run which is the recommended approach as it ensures all dependencies are properly managed by Poetry.

Dependencies

• pandas - Data manipulation and analysis
• numpy - Numerical computations
• matplotlib - Data visualization
• seaborn - Statistical data visualization
• scikit-learn - Machine learning algorithms
• jupyter - Interactive notebook environment

Project Structure

• credit_card_customer_segmentation.ipynb - Main Jupyter notebook containing the analysis
• customer_segmentation.csv - Dataset containing customer information
• data_dictionary.md - Detailed description of dataset features and their meanings
• dataset_documentation.ipynb - Additional documentation and exploratory analysis of the dataset
• README.md - Project documentation
• pyproject.toml & poetry.lock - Poetry dependency management configuration
• project_info.txt - Additional project notes and information

Documentation

For a detailed understanding of the dataset features and their meanings, please refer to:

1. data_dictionary.md - Contains definitions of all variables in the dataset
2. dataset_documentation.ipynb - Provides in-depth analysis and examples of the data structure

Analysis Steps

1. Data Loading and Exploration
2. Data Preprocessing and Feature Engineering
3. Feature Scaling
4. K-means Clustering
5. Cluster Analysis and Interpretation
6. Visualization of Results

Usage

To run the analysis:

1. Ensure all dependencies are installed:

poetry install

2. Launch Jupyter notebook:

poetry run jupyter notebook

3. Open credit_card_customer_segmentation.ipynb in the Jupyter interface