Choose from: overfitting, or underfitting: 1) If your model cannot even fit the training examples, then you have large bias. For this case, which fitting does it belong to: [Fill in the blank], ; 2) If you can fit the training data, but large error on testing data, then you probably have large variance. For this case, which fitting does it belong to: [Fill in the blank], ;