Problems

• Underfitting: The model is too simple to capture the underlying patterns in the data, leading to poor performance on both training and test data.
• Overfitting: The model is too complex and captures the biases in the training data, leading to poor generalization to new data.