FedAdagradServerOptimizer<T>
Models & Types · AiDotNet.FederatedLearning.ServerOptimizers
FedAdagrad server optimizer — adaptive learning rates using accumulated squared gradients.
Inheritance: FederatedLearningComponentBase<T> → FederatedServerOptimizerBase<T> → FedAdagradServerOptimizer<T>
For Beginners
Adagrad automatically reduces the learning rate for parameters that have been updated frequently, allowing rarely-updated parameters to learn faster. In federated settings, this helps handle heterogeneous client updates.
How It Works
Reference: Reddi, S., et al. (2021). “Adaptive Federated Optimization.” ICLR 2021.
Constructors
| Constructor | Summary |
|---|---|
FedAdagradServerOptimizer(double learningRate, double epsilon, double beta1) |
Methods
| Method | Summary |
|---|---|
Vector<T> Step(Vector<T> currentGlobalParameters, Vector<T> aggregatedTargetParameters) |
|
string GetOptimizerName() |