This code provides a method to represent categorical data by learned embedding vectors in the Gaussian Process regression of scikit-learn. Each level of the categorical variable is seen as an entity that is associated with a vector. The embedding vectors are learned from data.
Inside this folder run
pip install .
The file examples/MinimalExample.ipynb contains a minimal example.
The notebook requires some more dependencies:
In the folder examples run
pip install -r requirements.txt.
Then run jupyter lab