When input perturbation outperforms gradient perturbation: Achieving high-accuracy deep learning under local differential privacy.

@article{peng-2026-input, author = {Shunshun Peng and Jun Zhang and Chenxing Hu and Quanwang Wu and Hongbing Wang and Mengmeng Yang and Taolin Guo}, title = {{When input perturbation outperforms gradient perturbation: Achieving high-accuracy deep learning under local differential privacy}}, journal = {Information Processing and Management}, volume = {63}, number = {7}, pages = {104879}, year = {2026}, url = {https://doi.org/10.1016/J.IPM.2026.104879}, doi = {https://doi.org/10.1016/J.IPM.2026.104879} }