Classification Under Local Differential Privacy with Model Reversal and Model Averaging
{{output}}
Local differential privacy (LDP) has become a central topic in data privacy research, offering strong privacy guarantees by perturbing user data at the source and removing the need for a trusted curator. However, the noise introduced by LDP often significantly... ...