A supervised domain-agnostic framework that allows for probabilistic modelling, namely the prediction of probability distributions for individual data points.
The package offers a variety of features and specifically allows for
- the implementation of probabilistic prediction strategies in the supervised contexts
- comparison of frequentist and Bayesian prediction methods
- strategy optimization through hyperparamter tuning and ensemble methods (e.g. bagging)
- workflow automation
List of developers and contributors
The full documentation is available here.
Installation is easy using Python's package manager
$ pip install skpro
We welcome contributions to the skpro project. Please read our contribution guide.
If you use skpro in a scientific publication, we would appreciate citations.