GANDALFOptions<T>

Options & Configuration · AiDotNet.Models.Options

Configuration options for GANDALF (Gated Additive Neural Decision Forest).

Inheritance: ModelOptions → NeuralNetworkOptions → ModelHyperparameterOptions → FinancialNeuralNetworkOptions → RiskModelOptions<T> → GANDALFOptions<T>

For Beginners

GANDALF is like having a smart feature selector combined with a forest of decision trees.

Key ideas:

  1. Gated Feature Selection: Learns which features matter for each prediction
  2. Soft Decision Trees: Trees with smooth (differentiable) decisions
  3. Additive Ensemble: Trees are combined by adding their predictions

Why this works well:

  • Automatic feature importance learning (no manual selection needed)
  • Interpretable structure (can see which features and paths are used)
  • Combines benefits of neural networks and decision trees
  • Good handling of both numerical and categorical features

Example:

How It Works

GANDALF combines gated feature selection with neural decision trees in an additive ensemble. It learns which features are important through attention-based gating and makes predictions through soft decision trees.

Reference: “GANDALF: Gated Adaptive Network for Deep Automated Learning of Features” (2022)

Constructors

Constructor Summary
GANDALFOptions()

Properties

Property Summary
bool UseBatchNorm { get; set; } Gets or sets whether to use batch normalization.
double DropoutRate { get; set; } Gets or sets the dropout rate for regularization.
double InitScale { get; set; } Gets or sets the initialization scale for tree parameters.
double Temperature { get; set; } Gets or sets the temperature for soft tree decisions.
double WeightDecay { get; set; } Gets or sets the weight decay for regularization.
int GatingHiddenDimension { get; set; } Gets or sets the hidden dimension for gating network.
int LeafDimension { get; set; } Gets or sets the leaf dimension (output dimension per leaf).
int NumGatingLayers { get; set; } Gets or sets the number of gating layers.
int NumInternalNodes { get; } Gets the number of internal nodes per tree (2^depth - 1).
int NumLeaves { get; } Gets the number of leaves per tree (2^depth).
int NumTrees { get; set; } Gets or sets the number of trees in the ensemble.
int TreeDepth { get; set; } Gets or sets the depth of each tree.