Ordinal regression is suitable for predicting ordered outcomes like race positions. It accounts for the natural order in the target variable (e.g., 1st is better than 2nd, which is better than 3rd). You can use models that are specifically designed for ordinal outcomes.
Ordinal regression is suitable for predicting ordered outcomes like race positions. It accounts for the natural order in the target variable (e.g., 1st is better than 2nd, which is better than 3rd). You can use models that are specifically designed for ordinal outcomes.
https://pythonhosted.org/mord/
Your label will be the horse place (or, normalized place
0.0-1.0to handle the varying number of horses per race).