matetcho-steven/e-commerce-churn-ghana

# E-commerce Churn (Ghana) Data → RFM → Logistic Regression → Elasticsearch/Kibana Predict which customers are at risk of **churn** (no purchase in the last **45 days**) and visualize activity with **Elasticsearch + Kibana**. The notebook generates **synthetic, Ghana‑flavoured** data, builds **RFM** features, trains a **Logistic Regression** model, and ships event data to Elasticsearch for dashboards. --- ## Tech Stack - **Python**: `pandas`, `numpy`, `scikit-learn`, `matplotlib`, `seaborn`, `plotly`, `faker`, `duckdb` (optional) - **Model**: Logistic Regression (with `StandardScaler` in a pipeline) - **Big-data demo**: Elasticsearch 8.x + Kibana 8.x (via Docker) - **Notebook**: `E-commerce_insights_data_analysis.ipynb` --- ## Repository Structure ``` . ├─ E-commerce_insights_data_analysis.ipynb # end-to-end workflow (run top → bottom) ├─ docker-compose.yml # Elasticsearch + Kibana (Docker) ├─ requirements.txt # Python dependencies (pinned) ├─ data/ # generated CSVs (auto-created by notebook) └─ docs/ ├─ kibana_dashboard.png # add after you build dashboards ├─ roc_pr_curves.png # add after you plot metrics └─ lift_table.png # add after you export lift table ``` --- ## Prerequisites - Python **3.11+** - Docker Desktop (or Docker Engine) running - Windows users: PowerShell; macOS/Linux: Terminal --- ## Quickstart ### 1) Create & Activate a Python Environment ```bash # create project folder and enter mkdir ecommerce-churn-ghana && cd ecommerce-churn-ghana # create venv python -m venv .venv # activate it # Windows (PowerShell): . .venv/Scripts/Activate # macOS/Linux: source .venv/bin/activate ``` ### 2) Add Dependencies Create a file named `requirements.txt` with the following content: ```txt jupyterlab==4.2.5 pandas==2.2.2 numpy==1.26.4 scikit-learn==1.5.2 matplotlib==3.9.0 seaborn==0.13.2 plotly==5.24.1 faker …