Categorical data Machine Learning; Data Engineering
noun phrase
Definition: Data whose values belong to a finite set of discrete categories or classes rather than representing continuous numerical quantities [Wang et al. 2024].
Examples in context: “While several works follow the original ideas of transformers (vocabulary definition, pre-training, and fine-tuning) using categorical data as input, other proposals use numeric values as input.” [Siebra et al. 2024]
“Different choices are made for the dominant-probability category in a given quadrant, allowing generation of diverse categorical data.” [Wang et al. 2024]
Synonyms: qualitative data (partial overlap); category-valued data
Related terms: numerical data; tabular data; feature encoding; one-hot encoding; ordinal data