A phase in ML where the model has more parameters than the number of training samples, often leading to a high-variance, overfitted model.
Generality: 550
