Class CrossValidationFitDetectorOptions
Configuration options for detecting overfitting, underfitting, and good fitting in machine learning models using cross-validation techniques.
public class CrossValidationFitDetectorOptions- Inheritance
-
CrossValidationFitDetectorOptions
- Inherited Members
Remarks
The CrossValidationFitDetectorOptions class provides threshold settings that help determine whether a model is overfitting (performing well on training data but poorly on validation data), underfitting (performing poorly on both training and validation data), or has a good fit (performing well on both).
For Beginners: Think of these settings as the criteria for judging how well your AI model has learned. Just like when learning a new skill, an AI can learn too little (underfit), memorize without understanding (overfit), or learn just right (good fit). These thresholds help determine which category your model falls into by comparing how it performs on data it has seen before (training data) versus new data (validation data).
Properties
GoodFitThreshold
Gets or sets the threshold for identifying a good fit in a model.
public double GoodFitThreshold { get; set; }Property Value
- double
The good fit threshold as a decimal between 0 and 1, defaulting to 0.9 (90%).
Remarks
A good fit is identified when both training and validation performance exceed this threshold, and the difference between them is less than the overfit threshold.
For Beginners: This setting helps identify when your model has learned well. With the default value of 0.9, if your model performs above 90% of the maximum possible performance on both training and validation data, and the difference between them is small, it's considered to have a good fit. This is like a student who understands the material deeply and performs well on both practice tests and new questions, showing they've truly learned the subject rather than just memorizing answers.
OverfitThreshold
Gets or sets the threshold for detecting overfitting in a model.
public double OverfitThreshold { get; set; }Property Value
- double
The overfit threshold as a decimal between 0 and 1, defaulting to 0.1 (10%).
Remarks
Overfitting is detected when the difference between training performance and validation performance exceeds this threshold. A smaller value makes the detector more sensitive to overfitting.
For Beginners: This setting helps detect when your model is "memorizing" the training data instead of learning general patterns. With the default value of 0.1, if your model performs more than 10% better on training data than on validation data, it's considered to be overfitting. This is like a student who memorizes answers for a practice test but can't apply the knowledge to slightly different questions on the real test.
UnderfitThreshold
Gets or sets the threshold for detecting underfitting in a model.
public double UnderfitThreshold { get; set; }Property Value
- double
The underfit threshold as a decimal between 0 and 1, defaulting to 0.7 (70%).
Remarks
Underfitting is detected when the training performance is below this threshold. A higher value makes the detector more sensitive to underfitting.
For Beginners: This setting helps detect when your model hasn't learned enough from the training data. With the default value of 0.7, if your model's performance on training data is below 70% of the maximum possible performance, it's considered to be underfitting. This is like a student who hasn't studied enough and performs poorly even on practice questions they've seen before.