Definition
A straightforward hyperparameter optimization technique that systematically trains a model using every possible combination of a pre-defined subset of parameter values.
In short: A straightforward hyperparameter optimization technique that systematically trains a model using every possible combination of a pre-defined subset of parameter values.