Struct FederatedLearning 

pub struct FederatedLearning {
    pub config: FederatedConfig,
    pub clients: Vec<FederatedClient>,
    pub global_model: Arc<RwLock<ModelParameters>>,
    pub privacy_mechanism: PrivacyMechanism,
}

Federated Learning framework with privacy-preserving techniques

Implements federated averaging (FedAvg) and secure aggregation for privacy-preserving distributed machine learning.

Fields

config: FederatedConfig

Federated learning configuration

clients: Vec<FederatedClient>

Client models

global_model: Arc<RwLock<ModelParameters>>

Global model parameters

privacy_mechanism: PrivacyMechanism

Privacy mechanism

Implementations

impl FederatedLearning

pub fn new(config: FederatedConfig, num_features: usize) -> Self

Create a new federated learning system

pub fn add_client(&mut self, client_id: String, dataset_size: usize)

Add a client to the federated system

pub fn select_clients(&self) -> Vec<usize>

Select clients for a training round

pub fn federated_average( &self, client_updates: &[(usize, Vec<f64>)], ) -> Vec<f64>

Perform federated averaging of client updates

pub fn get_global_model(&self) -> ModelParameters

Get global model parameters

Trait Implementations

impl Debug for FederatedLearning

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.