ClusteredClientSelectionStrategy
Models & Types · AiDotNet.FederatedLearning.Selection
Cluster-based client selection using simple k-means over per-client embeddings.
Inheritance: ClientSelectionStrategyBase → ClusteredClientSelectionStrategy
How It Works
For Beginners: This strategy tries to pick clients from different “types” of behavior by clustering clients into groups and sampling from each cluster.
Constructors
| Constructor | Summary |
|---|---|
ClusteredClientSelectionStrategy(int clusterCount, int iterations) |
Methods
| Method | Summary |
|---|---|
List<int> SelectClients(ClientSelectionRequest request) |
|
string GetStrategyName() |