Random Search picks hyperparameter configurations randomly within given ranges. It’s often more efficient than exhaustive grid search, especially in high-dimensional spaces. Coupled with cross-validation, random search can be a practical way to find good hyperparameters quickly.