Skip to content

probabl-ai/calibration-cost-sensitive-learning

Repository files navigation

Probabilistic calibration and cost-sensitive learning

Browse the contents online

  • Launch JupyterLite Static HTML with precomputed results.

  • Launch JupyterLite In-browser interactive execution environment, no install needed.

Local setup

Install pixi

You can refer to the official website for installation.

Launching Jupyter Lab

To launch Jupyter Lab, run the following command:

pixi run jupyter lab

The Python environment and necessary packages will be automatically installed for you.

Opening lecture notes

The lecture notes are available in the content/python_files directory. To open the Python file as notebook, you need to right click on the file and select Open with -> Notebook.

Alternatively, you can generate notebooks as well:

pixi run -e doc convert-to-notebooks

This will convert the Python files into notebooks in the folder content/notebooks.

Building and testing the jupyterlite deployment locally

Test the deployment locally:

pixi run -e doc serve-jupyterlite

Then open http://[::]:8000/ in a web browser.

Note: firefox often complains about Service Workers not being initialized when running jupyterlite served from localhost. This prevents accessing the local dataset files. Use a different browser to test the local deployment.

About

Tutorial on probabilistic classification and cost-sensitive learning.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published