SESKStuff Everyone Should Know
SESK: Stuff Everyone Should Know · sesk.info/ai-underfitting-defined

A model does poorly even on its own training data. What is the likely problem?

Underfitting. The model has not captured the patterns in the data.

Start playing free

  1. Overfitting
  2. Data leakage
  3. Underfitting ✔

Why: An underfit model is too simple or undertrained, so it misses the real patterns. It scores badly on training data and new data alike.

Huh, didn't know that: Fixes include a more complex model, more useful features or training longer.

The dad joke
My underfit model predicted every house costs the same. Great for buyers, terrible for science.
Like this card?
Play the game

References

Answer: Underfitting. An underfit model is too simple or undertrained, so it misses the real patterns. It scores badly on training data and new data alike.

  1. Google Machine Learning Glossary: underfitting developers.google.com/machine-learning/glossary
    Producing a model with poor predictive ability because the model hasn't fully captured the complexity of the training data.
    Checked 2026-10-10.

Think this answer is wrong? Tap "Challenge this answer" on the card in the game and tell us why.

Test yourself in the game

Artificial IntelligenceA model scores great on its training data but badly on new data. What is that called?Artificial IntelligenceWhy document a dataset's intended use and context?Artificial IntelligenceHow do AI, machine learning and deep learning fit together?

That's stuff everyone should know.