Results for "patient risk"

A formal privacy framework ensuring outputs do not reveal much about any single individual’s data contribution.

A measure of a model class’s expressive capacity based on its ability to shatter datasets.

Measures a model’s ability to fit random noise; used to bound generalization error.