An Adaptive Communication-Efficient Federated Learning to Resist Gradient-Based Reconstruction Attacks

Table 1

Hyperparameter settings and letter representation.

Notation	Denote
	The number of all clients
	The number of compression
	The number of frequency
	Compression rate of single communication
	The fixed rounds of updating
	The mean parameters of all iterations’ communication
	The number of iterations when test accuracy is beyond 90 for the first time in all iterations
	The maximum of the accuracy in all iterations