Hyperparameter: Difference between revisions
imported>Unknown user No edit summary |
imported>Unknown user No edit summary |
||
(No difference)
| |||
Revision as of 02:53, 15 January 2026
Hyperparameter
In a differential privacy mechanism, a setting or parameter that controls a portion of the mechanism’s behavior or execution. The best setting may be data‐dependent, and a method that uses the confidential data as the basis for these parameters would not satisfy differential privacy. Examples include the clipping parameter for mechanisms that perform clipping, the number of iterations for iterative algorithms, and the learning rate or minibatch size for machine learning algorithms.
Source: NIST SP 800-226 | Category: