A model unable to produce sensible results on new data is also called “not able to generalize.” In this case, the model is too complex, and the patterns existing in the dataset are not well represented. Such a model with high variance overfits.

Overfitting models produce good predictions for data points in the training set but perform poorly on new samples.