Hyperparameters are external configurations chosen before a Machine Learning model begins its learning process. They guide how the model learns or its internal structure, distinct from Model Parameters which are learned from data. Their careful selection is crucial for a model's effectiveness.