juniorScikit-Learn
What is the fit-predict paradigm in Scikit-Learn?
Updated May 17, 2026
Short answer
It is the standard workflow where models are trained using fit() and used for predictions using predict().
Deep explanation
fit() learns patterns from training data, storing parameters. predict() applies learned parameters to unseen data for inference.
Real-world example
Used in classification problems like email spam detection.
Common mistakes
- Calling predict before fit or fitting on test data.
Follow-up questions
- What happens internally during fit()?
- Can predict be used on training data?