In a case study, if a linear regression model has a high R-squared value but poor predictive performance on new data, what might be the issue?
-
The model is too simple
-
The model is overfitting the training data
-
The model is underfitting the training data
-
The data is not linear
