Binning Machine Learning
noun
Definition: A discretization technique that divides continuous values into intervals, or bins, so that numeric features can be represented in a discrete form [scikit-learn Documentation].
Example in context: “The main idea is to plot the model’s accuracy as a function of its confidence. This is performed by binning predictions according to the associated predicted probability, and computing the accuracy for each bin.” [Lambert et al. 2022]
Synonyms: bucketing, discretization
Related terms: quantization, KBinsDiscretizer, histogram binning, ordinal encoding, one-hot encoding