What is Adjusted R-Squared
A statistical metric that evaluates the quality of a regression model, adjusted for the number of features.
Definition
The usual R-square shows how much of the variance in the data is explained by the model, but almost always increases as new features are added. Adjusted R-square takes into account the complexity of the model and helps to understand whether the new feature actually improves the explanation of the data.
Example
If the sales forecast gets a little better after adding dozens of random features, the adjusted R-squared may show that the improvement is not worth the complication.
Why it matters
The term helps avoid overestimating complex models and choosing understandable solutions, especially in analytics, finance and business forecasting.
How it works
The metric increases when a new feature actually improves the model, and can decrease if the feature adds complexity without benefit.
Where it is used
- regression analysis
- forecasting
- comparison of statistical models
Limitations
The metric does not replace testing on test data. A good value does not guarantee cause-and-effect relationships and robustness of the prediction.