Personal sandbox for rebuilding classic algorithms, exploring datasets, and sharing concise learning notes. Each subfolder contains either a standalone script or a Jupyter notebook that documents the experiment from data prep to evaluation.
- notebooks/ – interactive explorations (e.g., perceptron from scratch, PCA intuition).
- datasets/ – small, open datasets kept alongside experiments for reproducibility.
- pyproject.toml – dependency and tooling metadata for reproducible environments.
Expect lightweight implementations with clear commentary, comparisons against baseline models, and incremental refactors as I iterate on ideas.
- Create and activate a Python 3.11+ virtual environment (e.g.,
uv venv && source .venv/bin/activate). - Install dependencies:
uv pip install -e .(editable install keeps notebooks aligned with the latest utilities). - Launch Jupyter Lab or VS Code notebooks to run the explorations.
Tip: The notebooks are written to be rerun end-to-end; they avoid hidden state and set random seeds where relevant.
This repository primarily supports my learning journey, but feedback is welcome via GitHub issues or pull requests. If you re-run or extend an experiment, mention the original notebook and describe the change so others can follow along.
Happy experimenting!