overfitting

Overfitting is a scenario in ML where the model has sort of memorized the data and its patterns.

This typically occurs if the model has more parameters than can be justified for the dataset. In such cases, the model becomes too sensitive to the data and thus may not generalize well to future observation points.