TimeSeriesRegressionOptions<T>
Options & Configuration · AiDotNet.Models.Options
Configuration options for time series regression models, which analyze data collected over time to identify patterns and make predictions.
Inheritance: ModelOptions → RegressionOptions<T> → TimeSeriesRegressionOptions<T>
Known derived types: ARIMAOptions<T>, ARIMAXModelOptions<T>, ARMAOptions<T>, ARModelOptions<T>, AutoformerOptions<T>, BayesianStructuralTimeSeriesOptions<T>, CCDMOptions<T>, CSDIOptions<T>, ChronosBoltOptions<T>, ChronosOptions<T>, DCRNNOptions<T>, DLinearOptions<T>, DeepANTOptions<T>, DeepAROptions<T>, DeepFactorOptions<T>, DeepStateOptions<T>, DiffusionTSOptions<T>, DynamicRegressionWithARIMAErrorsOptions<T>, ExponentialSmoothingOptions<T>, FlowStateOptions<T>, GARCHModelOptions<T>, GPT4TSOptions<T>, GraphWaveNetOptions<T>, HippoOptions<T>, InformerOptions<T>, InterventionAnalysisOptions<T, TInput, TOutput>, KairosOptions<T>, KronosOptions<T>, LLMTimeOptions<T>, LSTMVAEOptions<T>, LSTNetOptions<T>, LagLlamaOptions<T>, MAModelOptions<T>, MGTSDOptions<T>, MOIRAIOptions<T>, MOMENTOptions<T>, MQCNNOptions<T>, MTGNNOptions<T>, Mamba2Options<T>, MambaOptions<T>, NBEATSModelOptions<T>, NHiTSOptions<T>, NLinearOptions<T>, NeuralNetworkARIMAOptions<T>, ProphetOptions<T, TInput, TOutput>, RWKVForecastingOptions<T>, RelationalGCNOptions<T>, S4Options<T>, SARIMAOptions<T>, STGNNOptions<T>, STLDecompositionOptions<T>, ScoreGradOptions<T>, SimMTMOptions<T>, SpectralAnalysisOptions<T>, StateSpaceModelOptions<T>, SundialOptions<T>, TBATSModelOptions<T>, TCNOptions<T>, TESTOptions<T>, TFCOptions<T> (+26 more)
For Beginners
Time series regression helps you analyze and predict data that changes over time, like stock prices, weather patterns, or monthly sales figures. Unlike regular regression that just looks for relationships between variables, time series regression also considers when things happened. It can detect patterns like “sales always increase in December” or “temperature today is related to temperature yesterday.” This class lets you configure how the model analyzes these time-based patterns.
How It Works
Time series regression extends traditional regression analysis to account for the temporal nature of data, where observations are collected sequentially over time. These models can capture trends, seasonal patterns, and the effects of past values on current and future values.
Constructors
| Constructor | Summary |
|---|---|
TimeSeriesRegressionOptions() |
Initializes a new instance with default settings. |
TimeSeriesRegressionOptions(TimeSeriesRegressionOptions<T> other) |
Creates a deep copy of an existing TimeSeriesRegressionOptions, including the early-stopping settings that a shallow clone would otherwise reset to their defaults. |
Properties
| Property | Summary |
|---|---|
ILossFunction<T> LossFunction { get; set; } |
Gets or sets the loss function used for gradient computation and model training. |
TimeSeriesModelType ModelType { get; set; } |
Gets or sets the specific type of time series model to use. |
bool AutocorrelationCorrection { get; set; } |
Gets or sets whether to apply autocorrelation correction to the model. |
bool IncludeTrend { get; set; } |
Gets or sets whether to include a trend component in the model. |
bool UseEarlyStopping { get; set; } |
Gets or sets whether training stops once the monitored loss stops improving. |
double EarlyStoppingMinDelta { get; set; } |
Smallest improvement in validation loss that counts as progress for early stopping. |
double? MaxPredictionAbsValue { get; set; } |
Gets or sets the maximum absolute value allowed for predictions. |
int EarlyStoppingPatience { get; set; } |
Number of epochs without improvement after which training stops early. |
int LagOrder { get; set; } |
Gets or sets the lag order, which determines how many previous time steps are used as predictors. |
int MaxTrainingTimeSeconds { get; set; } |
Wall-clock training budget in seconds. |
int NumFeatures { get; set; } |
Gets or sets the number of input features (variables) observed at each time step. |
int SeasonalPeriod { get; set; } |
Gets or sets the seasonal period of the time series data. |