CTABGANPlusOptions<T>

Options & Configuration · AiDotNet.Models.Options

Configuration options for CTAB-GAN+, an enhanced conditional tabular GAN with auxiliary classifier discriminator and mixed-type encoder for high-quality synthetic data generation.

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

For Beginners

CTAB-GAN+ is an improved version of CTGAN that produces higher-quality synthetic data by using smarter training signals:

  1. Better discriminator: Not only decides real/fake, but also classifies data categories
  2. Better encoding: Handles rare categories better with log-frequency encoding
  3. Better evaluation: Checks if generated data is useful for downstream ML tasks

Example:

How It Works

CTAB-GAN+ extends CTGAN with several architectural improvements:

  • Auxiliary Classifier GAN (ACGAN): Discriminator also predicts class labels
  • Mixed-type encoder: Log-frequency encoding for long-tail categoricals
  • Downstream losses: Additional classification/regression losses for utility
  • Conditional vector: Same training-by-sampling as CTGAN

Reference: “CTAB-GAN: Effective Table Data Synthesizing” (Zhao et al., 2021)

Constructors

Constructor Summary
CTABGANPlusOptions() Initializes a new instance with the default settings.
CTABGANPlusOptions(CTABGANPlusOptions<T> other) Initializes a new instance by copying every setting of another instance.

Properties

Property Summary
double ClassifierWeight { get; set; } Gets or sets the weight for the auxiliary classifier loss in the discriminator.
double DiscriminatorDropout { get; set; } Gets or sets the dropout rate for the discriminator.
double GradientPenaltyWeight { get; set; } Gets or sets the gradient penalty weight for WGAN-GP.
double GumbelTemperature { get; set; } Gets or sets the Gumbel-softmax temperature of the generator’s mode and discrete-column heads.
double InformationWeight { get; set; } Gets or sets the weight for the information loss (statistical similarity).
double LearningRate { get; set; } Gets or sets the learning rate for both generator and discriminator.
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 fed to the generator.
int Epochs { get; set; } Gets or sets the number of training epochs.
int PacSize { get; set; } Gets or sets the PacGAN packing size.
int TargetColumnIndex { get; set; } Gets or sets the index of the target/label column for the auxiliary classifier.
int VGMModes { get; set; } Gets or sets the number of VGM modes for continuous column normalization.
int[] DiscriminatorDimensions { get; set; } Gets or sets the hidden layer sizes for the discriminator network.
int[] GeneratorDimensions { get; set; } Gets or sets the hidden layer sizes for the generator network.