← Library · Core concept
Model Calibration
Model calibration refers to how well a model's predicted probabilities match the true likelihood of an event occurring. A well-calibrated model, for example, should predict a 70% chance of rain, and it should actually rain about 70% of the times it makes that prediction. Miscalibration means the model is overconfident or underconfident in its predictions.
In plain terms
It's like a weather forecaster whose prediction of '70% chance of rain' actually means it rains 7 out of 10 times they say it.
Why it matters
Calibrated models are essential for decision-making in high-stakes environments where knowing the true probability of an outcome is critical, not just the predicted outcome itself.
Learn one new AI thing every day.
Daily Deck sends you seven plain-English cards like this every morning. Free.
Start free