ResidualAnalysisFitDetector<T, TInput, TOutput>
Models & Types · AiDotNet.FitDetectors
A detector that evaluates model fit quality by analyzing the residuals (errors) of the model.
Inheritance: FitDetectorBase<T, TInput, TOutput> → ResidualAnalysisFitDetector<T, TInput, TOutput>
For Beginners
This class helps you understand how well your model is performing by looking at the “residuals” - the differences between what your model predicted and the actual values.
Think of residuals like the errors your model makes. By analyzing these errors in different ways, we can tell if your model:
- Is generally accurate (good fit)
- Consistently makes errors in the same direction (bias)
- Makes wildly different errors each time (high variance)
- Works well on training data but poorly on new data (overfitting)
- Doesn’t capture the complexity of your data (underfitting)
This detector examines these patterns to give you a clear picture of your model’s performance and how to improve it.
Constructors
| Constructor | Summary |
|---|---|
ResidualAnalysisFitDetector(ResidualAnalysisFitDetectorOptions options) |
Initializes a new instance of the ResidualAnalysisFitDetector class. |
Methods
| Method | Summary |
|---|---|
FitDetectorResult<T> DetectFit(ModelEvaluationData<T, TInput, TOutput> evaluationData) |
Analyzes the model’s performance data and determines the quality of fit based on residual analysis. |