Table of Contents

Interface ITimeSeriesModel<T>

Namespace
AiDotNet.Interfaces
Assembly
AiDotNet.dll

Defines the core functionality for time series prediction models.

public interface ITimeSeriesModel<T> : IFullModel<T, Matrix<T>, Vector<T>>, IModel<Matrix<T>, Vector<T>, ModelMetadata<T>>, IModelSerializer, ICheckpointableModel, IParameterizable<T, Matrix<T>, Vector<T>>, IFeatureAware, IFeatureImportance<T>, ICloneable<IFullModel<T, Matrix<T>, Vector<T>>>, IGradientComputable<T, Matrix<T>, Vector<T>>, IJitCompilable<T>

Type Parameters

T

The numeric data type used for calculations (e.g., float, double).

Inherited Members
Extension Methods

Remarks

Time series models analyze sequential data points collected over time to identify patterns and make predictions about future values.

For Beginners: A time series model helps you predict future values based on past data that was collected in sequence over time. For example:

  • Predicting tomorrow's temperature based on weather patterns from the past week
  • Forecasting next month's sales based on previous months' sales data
  • Estimating website traffic for next week based on historical visitor counts

These models look for patterns in your historical data (like trends, seasonal effects, and cycles) to make educated guesses about what will happen next.

This interface inherits from IModelSerializer, which means these models can be saved to disk and loaded back later - useful for when you've trained a good model and want to use it again without retraining.

Methods

EvaluateModel(Matrix<T>, Vector<T>)

Evaluates the model's performance using test data.

Dictionary<string, T> EvaluateModel(Matrix<T> xTest, Vector<T> yTest)

Parameters

xTest Matrix<T>

The matrix of input features for testing.

yTest Vector<T>

The vector of actual target values corresponding to xTest.

Returns

Dictionary<string, T>

A dictionary of evaluation metrics with metric names as keys and metric values as values.

Remarks

This method measures how well the model's predictions match actual values on data it hasn't seen during training.

For Beginners: This method helps you check how accurate your model is. You provide:

  • xTest: Input data that the model hasn't seen during training
  • yTest: The actual correct values for those inputs

The method will:

  1. Use your model to make predictions for xTest
  2. Compare those predictions to the actual values in yTest
  3. Calculate various error metrics to tell you how close the predictions were

Common metrics include:

  • Mean Absolute Error (MAE): The average size of the errors
  • Mean Squared Error (MSE): Similar to MAE but penalizes large errors more
  • R-squared: How much of the variation in the data is explained by the model

Lower error values and higher R-squared values generally indicate a better model.

PredictSingle(Vector<T>)

Predicts a single value based on the provided input vector.

T PredictSingle(Vector<T> input)

Parameters

input Vector<T>

The input vector containing features for prediction.

Returns

T

The predicted value for the given input.

Remarks

This method generates a single prediction based on the input vector, providing a convenient way to get individual predictions without creating a matrix.

For Beginners: This lets you get a prediction for a single point in time rather than a whole series of predictions at once. It's like asking "What's the forecast for tomorrow?" instead of "What's the forecast for the next week?"

Edit this page