When a model learns the training data too closely and performs poorly on sufficiently different examples. Assess it using data not used for training and representative real-world conditions.
When a model learns the training data too closely and performs poorly on sufficiently different examples. Assess it using data not used for training and representative real-world conditions.