TableGANOptions<T>

Options & Configuration · AiDotNet.Models.Options

Configuration options for TableGAN, a DCGAN-style generative adversarial network for synthesizing tabular data with classification and information loss regularization.

Inheritance: ModelOptions → NeuralNetworkOptions → ModelHyperparameterOptions → FinancialNeuralNetworkOptions → RiskModelOptions<T> → SyntheticDataGeneratorOptions<T> → TableGANOptions<T>

For Beginners

TableGAN generates fake tabular data using three quality checks:

  1. “Does it look real?” — the discriminator judges overall realism
  2. “Are the labels correct?” — a classifier checks that labels match the data
  3. “Are the statistics right?” — mean/variance of synthetic data should match the real data

This triple-loss approach produces higher quality synthetic data than a standard GAN alone.

Example:

How It Works

TableGAN uses a DCGAN architecture adapted for tabular data with three loss components:

  • Adversarial loss: Standard GAN objective for realism
  • Classification loss: Ensures generated data preserves label column relationships
  • Information loss: Minimizes statistical divergence between real and synthetic data

Reference: “Data Synthesis based on Generative Adversarial Networks” (Park et al., 2018)

Constructors

Constructor Summary
TableGANOptions()

Properties

Property Summary
bool TrainClassifier { get; set; } Gets or sets whether the generator’s loss includes the classification auxiliary term.
double ClassificationWeight { get; set; } Gets or sets the weight of the classification loss.
double DiscriminatorDropout { get; set; } Gets or sets the dropout rate for discriminator hidden layers.
double GradientPenaltyWeight { get; set; } Gets or sets the weight for the WGAN-GP gradient penalty term.
double InformationWeight { get; set; } Gets or sets the weight of the information loss (statistical similarity).
double LearningRate { get; set; } Gets or sets the learning rate.
int BatchSize { get; set; } Gets or sets the training batch size.
int DiscriminatorSteps { get; set; } Gets or sets the number of discriminator training steps per generator step.
int EmbeddingDimension { get; set; } Gets or sets the dimension of the random noise vector for the generator.
int Epochs { get; set; } Gets or sets the number of training epochs.
int LabelColumnIndex { get; set; } Gets or sets the index of the label column (for classification loss).
int VGMModes { get; set; } Gets or sets the number of VGM modes for continuous column transformation.
int[] ClassifierDimensions { get; set; } Gets or sets the hidden layer sizes for the classifier head.
int[] DiscriminatorDimensions { get; set; } Gets or sets the hidden layer sizes for the discriminator.
int[] GeneratorDimensions { get; set; } Gets or sets the hidden layer sizes for the generator.