Data variability Machine Learning
noun phrase
Definition: The extent to which data values differ from one another or are spread around a location value such as the mean; in statistical terminology, this corresponds to the spread or variability of a dataset. NIST explicitly defines variability as the spread of observed values about some location value [NIST].
Example in context: “For few-sample experiments, we fixed the same data subset across all models to avoid performance changes related to data variability.” [Razdaibiedina et al. 2023]
Synonyms: variability (in statistical context); dispersion; spread
Related terms: variance, standard deviation