TimeSeriesCrossValidationFitDetector<T, TInput, TOutput>

Models & Types · AiDotNet.FitDetectors

A specialized detector that evaluates how well a model fits time series data using cross-validation techniques.

Inheritance: FitDetectorBase<T, TInput, TOutput> → TimeSeriesCrossValidationFitDetector<T, TInput, TOutput>

For Beginners

This class helps you understand if your time series model is learning correctly from your data.

Time series data is information collected over time in sequence (like daily temperatures, monthly sales, or hourly website traffic). When working with time series data, we need special techniques because:

  1. The order of data matters (unlike regular data where order doesn’t matter)
  2. Recent data is often more important than older data
  3. There might be patterns that repeat over time (like seasonal patterns)

This detector analyzes your model’s performance and tells you if it’s:

  • Learning too much detail from your data (overfitting)
  • Not learning enough patterns (underfitting)
  • Performing inconsistently (high variance)
  • Working well (good fit)
  • Behaving unpredictably (unstable)

It also provides specific recommendations to improve your model based on these findings.

Constructors

Constructor Summary
TimeSeriesCrossValidationFitDetector(TimeSeriesCrossValidationFitDetectorOptions options) Initializes a new instance of the TimeSeriesCrossValidationFitDetector class.

Methods

Method Summary
FitDetectorResult<T> DetectFit(ModelEvaluationData<T, TInput, TOutput> evaluationData) Analyzes model performance data and determines how well the model fits the time series data.