TS2VecOptions<T>
Options & Configuration · AiDotNet.Models.Options
Configuration options for TS2Vec (Contrastive Learning of Universal Time Series Representations).
Inheritance: ModelOptions → RegressionOptions<T> → TimeSeriesRegressionOptions<T> → TS2VecOptions<T>
How It Works
TS2Vec learns universal time series representations via hierarchical contrastive learning across augmented context views, producing contextual representations at arbitrary granularities.
Reference: Yue et al., “TS2Vec: Towards Universal Representation of Time Series”, AAAI 2022.
Constructors
| Constructor | Summary |
|---|---|
TS2VecOptions() |
Creates options with the default values. |
TS2VecOptions(TS2VecOptions<T> other) |
Creates a copy of another options instance. |
Properties
| Property | Summary |
|---|---|
double DropoutRate { get; set; } |
Dropout probability applied during training. |
double InstanceContrastiveWeight { get; set; } |
Weight of the instance-wise contrastive loss. |
double TemporalContrastiveWeight { get; set; } |
Weight of the temporal contrastive loss. |
int ContextLength { get; set; } |
Number of past time steps the model reads as context. |
int ForecastHorizon { get; set; } |
Number of future time steps to forecast. |
int HiddenDimension { get; set; } |
Width of the encoder’s hidden representations. |
int NumLayers { get; set; } |
Number of dilated convolution blocks of the encoder. |
int OutputDimension { get; set; } |
Size of each time step’s output representation. |