ashrafutuyubahe/rwanda-vehicle-prediction

# Vehicle ML Lab -- Django Machine Learning Project A Django web app that performs regression, classification, and clustering on a vehicle sales dataset from Rwanda (1000 records, 30 districts). Built for the ML lab exercise. Covers EDA, price prediction, income classification, client segmentation, and a Plotly map of Rwanda districts. ## Project structure ``` Django_ml_lab/ vehicles_ml_dataset.csv = 0.70) refines the silhouette score to 0.94. Coefficient of variation is also calculated. ## Exercise answers **(a)** Rwanda district map with boundaries and client counts per district -- rendered on the EDA page using Plotly scattermapbox with bubble markers and boundary polygons for all 30 districts. **(b)** Coefficient of variation and silhouette score are displayed on the clustering page. Silhouette refined above 0.9 using PowerTransformer + core sample filtering (no re-clustering). ## Tech stack - Django 5.2 - scikit-learn (RandomForest, KMeans, PowerTransformer) - pandas, numpy - Plotly.js for the Rwanda map - Bootstrap 5.3 for the frontend - joblib for model serialization