What Is Underfitting?

What is Underfitting?

Underfitting happens when a model is too simple or not trained enough, causing it to fail at learning important patterns in the data. As a result, it performs poorly on both training and new data.

This is like trying to learn a language using only five words—you don’t know enough to understand or respond correctly.

Why Underfitting Matters

• Poor training accuracy: The model cannot even understand the data it sees.
• Bad real-world results: Predictions remain low-quality and unreliable.
• Too-simple model structure: Essential features are ignored.

Signs of Underfitting

• Low accuracy on both training and test data.
• Very small model architecture lacking layers or complexity.
• Insufficient training time or not enough epochs.

How to Fix Underfitting

• Increase model complexity (more layers or parameters).
• Train longer to allow deeper learning.
• Use more meaningful features or better training data.
• Reduce regularization so the model can learn more details.