TimeMAEOptions<T>
Options & Configuration · AiDotNet.Models.Options
Configuration options for TimeMAE (Masked Autoencoder for Time Series).
Inheritance: ModelOptions → RegressionOptions<T> → TimeSeriesRegressionOptions<T> → TimeMAEOptions<T>
How It Works
TimeMAE applies masked autoencoding to time series, randomly masking patches of the input and training a transformer to reconstruct the missing patches, learning rich temporal representations.
Reference: Cheng et al., “TimeMAE: Self-Supervised Representations of Time Series with Decoupled Masked Autoencoders”, 2023.
Constructors
| Constructor | Summary |
|---|---|
TimeMAEOptions() |
Creates options with the default values. |
TimeMAEOptions(TimeMAEOptions<T> other) |
Creates a copy of another options instance. |
Properties
| Property | Summary |
|---|---|
double DropoutRate { get; set; } |
Dropout probability applied during training. |
double LearningRate { get; set; } |
Gets or sets the Adam learning rate used for native training. |
double MaskRatio { get; set; } |
Fraction of patches masked in each training series. |
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 model’s hidden representations. |
int NumDecoderLayers { get; set; } |
Number of decoder layers. |
int NumEncoderLayers { get; set; } |
Number of encoder layers. |
int NumHeads { get; set; } |
Number of attention heads. |
int PatchLength { get; set; } |
Number of time steps in each patch the series is split into. |