INO-SGD: Addressing Utility Imbalance under Individualized Differential Privacy
INO-SGD:在个性化差分隐私下解决效用不平衡问题
机构 * Department of Computer Science, National University of Singapore(新加坡国立大学计算机科学系) ; Agency for Science, Technology and Research (A*STAR), Singapore(新加坡科技研究局)
AI总结 本文提出INO-SGD算法,通过在每个批次中策略性降低权重,改善所有迭代中更隐私数据的性能,满足个性化差分隐私需求,解决现有方法无法兼顾隐私和效用的不足。
Comments Accepted to the 14th International Conference on Learning Representations (ICLR-26)