vault backup: 2026-08-11 17:53:14

Wiki/Machine Learning/Imbalanced Data.md

@@ -22,3 +22,5 @@ At the level of the algorithm there are a few approaches that can be taken to mi
 - Some models can weight different classes to prioritize one over another
 - The choice of model to use is important in handling imbalanced data
 	- Ensemble models such as random forests are good at handling imbalanced data
+
+Algorithm-level approaches are most commonly the choice of the model used and weighting doesn't work with every model and can be challenging to get working. Knowing what model to use for different situations (especially when data is imbalanced) is very useful to get the best performance on a dataset.