LSTNetOptions<T>
Options & Configuration · AiDotNet.Models.Options
Configuration options for the LSTNet (Long Short-Term Time-series Network) model.
Inheritance: ModelOptions → RegressionOptions<T> → TimeSeriesRegressionOptions<T> → LSTNetOptions<T>
For Beginners
LSTNet is like having multiple specialists working together to predict the future:
- The convolutional part is like scanning for local patterns - like noticing “sales always spike on weekends”
- The recurrent part remembers long-term trends - like “sales grow 10% each month”
- The skip-RNN looks for seasonal patterns - like “Christmas sales are always highest”
- The autoregressive part handles simple linear trends - like “each day is slightly higher than yesterday”
By combining all these, LSTNet can capture complex patterns in data where multiple time scales matter, such as electricity consumption (hourly, daily, weekly patterns), stock prices, or traffic flow.
How It Works
LSTNet is a neural network architecture specifically designed for multivariate time series forecasting. It combines multiple components to capture patterns at different temporal scales:
- Convolutional layers for short-term local patterns
- Recurrent layers (GRU) for long-term dependencies
- Skip-RNN for very long periodic patterns
- Autoregressive component for local linear trends
Constructors
| Constructor | Summary |
|---|---|
LSTNetOptions() |
Initializes a new instance of the LSTNetOptions class with default values. |
LSTNetOptions(LSTNetOptions<T> other) |
Initializes a new instance by copying from another instance. |
Properties
| Property | Summary |
|---|---|
bool UseHighway { get; set; } |
Gets or sets whether to use highway connections. |
double DropoutRate { get; set; } |
Gets or sets the dropout rate for regularization. |
double LearningRate { get; set; } |
Gets or sets the learning rate for training. |
int AutoregressiveWindow { get; set; } |
Gets or sets the window size for the autoregressive component. |
int BatchSize { get; set; } |
Gets or sets the batch size for training. |
int ConvolutionFilters { get; set; } |
Gets or sets the number of convolutional filters. |
int ConvolutionKernelSize { get; set; } |
Gets or sets the kernel size for convolutional layers. |
int Epochs { get; set; } |
Gets or sets the number of training epochs. |
int ForecastHorizon { get; set; } |
Gets or sets the forecast horizon (prediction length). |
int HiddenRecurrentSize { get; set; } |
Gets or sets the hidden state size of the main recurrent layers (GRU/LSTM). |
int HiddenSkipSize { get; set; } |
Gets or sets the hidden state size of the skip recurrent layers. |
int LookbackWindow { get; set; } |
Gets or sets the lookback window size (context length). |
int SkipPeriod { get; set; } |
Gets or sets the skip period for the Skip-RNN component. |
int SkipRecurrentLayers { get; set; } |
Gets or sets the number of skip recurrent layers. |