The distance between M and F is sqrt((0.5)^2) = 0.5. In some experiments, the plain one-over-n-hot encoding for binary categorical data did not work as well as the attenuated encoding. But results ...

Categorical vs. Continuous Data: Real-World Scenarios Choosing whether to present data in categories or according to quantitative value depends on what you want to accomplish.

Decision tree models can handle both numerical and categorical theoretically, but it depends on how the models are implemented in practice; for example, scikit-learn uses an optimized version of ...