Class FederatedClientDataset<TInput, TOutput>
Represents a single client's local dataset for federated learning.
public sealed class FederatedClientDataset<TInput, TOutput>
Type Parameters
TInputThe type of the input data (e.g., Matrix<T> or Tensor<T>).
TOutputThe type of the output data (e.g., Vector<T> or Tensor<T>).
- Inheritance
-
FederatedClientDataset<TInput, TOutput>
- Inherited Members
Remarks
This type is intentionally simple: it holds the client's local features and labels plus the sample count used for weighting during aggregation.
Constructors
FederatedClientDataset(TInput, TOutput, int)
Initializes a new instance of the FederatedClientDataset<TInput, TOutput> class.
public FederatedClientDataset(TInput features, TOutput labels, int sampleCount)
Parameters
featuresTInputThe client's local feature data.
labelsTOutputThe client's local label/target data.
sampleCountintThe number of samples contained in this dataset.
Properties
Features
Gets the client's local feature data.
public TInput Features { get; }
Property Value
- TInput
Labels
Gets the client's local label/target data.
public TOutput Labels { get; }
Property Value
- TOutput
SampleCount
Gets the number of samples contained in this dataset.
public int SampleCount { get; }