AutoML Benchmarking

This project compares the performance of different AutoML frameworks (FLAML, Auto-sklearn) on standard datasets.

Structure

• automl_tools/: Scripts to run individual AutoML tools.
• notebooks/: Jupyter notebooks for analysis and visualization.
• data/: Datasets used.
• results/: Saved results and plots.
• requirements/: Dependency groups for production and development.
• Dockerfile & docker-compose.yml: Containerized development environment.

Setup (with uv)

uv venv
source .venv/bin/activate
uv pip install -r requirements/dev.txt

Run (Docker)

docker-compose up --build

Usage

Run the notebook in notebooks/ to execute and analyze the benchmarks.