Documentation site for the House-Price-Prediction project — a multi-model ML system that predicts residential property sale prices from structural and quality features.
gozieibekwe.github.io/house-price-docs
This is a standalone documentation project, separate from the source code repository. It follows a docs-as-code approach:
- Documentation lives in Markdown files under
docs/ mkdocs.ymldefines the site structure and theme- GitHub Actions automatically builds and deploys the site to GitHub Pages on every push to
main
house-price-docs/
├── docs/
│ ├── index.md # Project overview
│ ├── getting-started.md # Installation and setup
│ ├── methodology.md # Dataset, features, model training
│ ├── app-reference.md # Input fields and output documentation
│ └── results.md # Model comparison and selection rationale
├── .github/
│ └── workflows/
│ └── deploy.yml # CI/CD: auto-deploy to GitHub Pages
├── mkdocs.yml # MkDocs configuration
└── README.md
pip install mkdocs-material
mkdocs serveOpens a live-reloading preview at http://localhost:8000. Changes to any file under docs/ update the preview instantly.
Pushing to main triggers the GitHub Actions workflow in .github/workflows/deploy.yml, which builds the site and deploys it to the gh-pages branch. GitHub Pages serves the built site automatically.
No manual build step required.