Dennisogaka/Japan-Car-Import-Advisory-Platform

An end-to-end machine learning pipeline that scrapes automotive listings, applies rigorous data cleaning, and trains a unified scikit-learn model to predict vehicle values, integrating local port costs and KRA (Kenya Revenue Authority) tax charges to calculate total import costs and deliver a direct financial comparison against purchasing locally. # Vehicle Price Prediction and Comparison Pipeline This is a Machine Learning pipeline that scrapes vehicle listings, cleans the data, and trains a scikit-learn model to predict values, integrating KRA tax schedules and port fees to calculate total landed import costs and compare them directly against local market prices. --- ## Project Lifecycle Workflow 1. **Data Scraping**: Extracting raw vehicle listings and specifications. 2. **Data Cleaning**: Handling missing values, removing duplicates, and fixing data types. 3. **Model Training**: Engineering features and training a machine learning pipeline (`ColumnTransformer` + Regression). 4. **Deployment**: Serving the model locally via Flask API endpoint. --- ## Project Structure ```text ├── notebook/ │ ├── 1_japan_car.ipynb ├── car_import_advisory_api.py ├── car_price_prediction.py ├── car_pricing_model_V1.pkl └── README.md