FedC2I framework enables adaptive parameter aggregation in Federated Learning

FedC2I introduces a method to quantify client-level and class-level influence within federated learning systems. This allows for adaptive parameter aggregation by leveraging the mutual influence between clients with similar preferences, rather than relying on fixed weighting.