Overfitting
Definition
Overfitting adapts a model too strongly to its training data.
A state in which a model fits training data including random patterns too closely and generalizes worse on new data.
Simply explained
The model also learns random peculiarities of its training examples and performs worse with new examples.
Why it matters in practice
The term helps to select, develop, evaluate AI systems in a well-founded manner and realistically assess their technical risks.
Example from business
A complex model achieves nearly perfect training scores but performs significantly worse on temporally separated tests.
Limits and pitfalls
- Only consider training errors or repeatedly use test data for model selection.
Origin & history
Overfitting is a fundamental problem of statistical modeling and machine learning.
Use in mid-sized companies
- Possible example: A complex model achieves nearly perfect training scores but performs significantly worse on temporally separated tests.
Frequently asked questions
What does overfitting mean?
A state in which a model fits training data including random patterns too closely and generalizes worse on new data. What mistakes should be avoided? Only consider training errors or repeatedly use test data for model selection.
Sources
Source checked on 10.10.2026. Editorial application examples are hypothetical; no proof of real customer results.
Editorially reviewed on: