IFederatedUnlearner<T>
Interfaces · AiDotNet.FederatedLearning.Unlearning
Core interface for federated unlearning: removes a client’s contribution from the global model.
Known implementations: DiffusiveNoiseUnlearner<T>, ExactRetrainingUnlearner<T>, GradientAscentUnlearner<T>, InfluenceFunctionUnlearner<T>
For Beginners
This interface defines the “forget” operation for federated learning.
When a client exercises their GDPR right to be forgotten, you call Unlearn with
their client ID. The unlearner modifies the global model to remove that client’s influence
and returns a certificate proving it was done.
How It Works
Usage:
Properties
| Property | Summary |
|---|---|
string MethodName { get; } |
Gets the name of this unlearning method. |
Methods
| Method | Summary |
|---|---|
ValueTuple<Tensor<T>, UnlearningCertificate> Unlearn(int targetClientId, Tensor<T> globalModel, Dictionary<int, List<Tensor<T>>> clientHistories) |
Removes a client’s contribution from the global model. |