SyntheticTabularFederatedBenchmarkOptions
Options & Configuration · AiDotNet.Configuration
Configuration options for the synthetic federated tabular benchmark suite.
For Beginners
Synthetic means AiDotNet generates the data automatically instead of reading files from disk. This is useful for CI and quick sanity checks.
How It Works
This suite generates a deterministic synthetic dataset with non-IID client distributions, allowing benchmark runs without external dataset dependencies.
Constructors
| Constructor | Summary |
|---|---|
SyntheticTabularFederatedBenchmarkOptions() |
Properties
| Property | Summary |
|---|---|
SyntheticTabularTaskType TaskType { get; set; } |
Gets or sets the synthetic task type. |
double DirichletAlpha { get; set; } |
Gets or sets the Dirichlet concentration parameter controlling label skew across clients. |
double NoiseStdDev { get; set; } |
Gets or sets the noise standard deviation applied to generated targets/scores. |
int? ClassCount { get; set; } |
Gets or sets the number of classes for classification tasks (null uses defaults). |
int? ClientCount { get; set; } |
Gets or sets the number of federated clients to simulate (null uses defaults). |
int? FeatureCount { get; set; } |
Gets or sets the number of input features per sample (null uses defaults). |
int? TestSamplesPerClient { get; set; } |
Gets or sets the number of test samples per client (null uses defaults). |
int? TrainSamplesPerClient { get; set; } |
Gets or sets the number of training samples per client (null uses defaults). |