信息安全研究2024，Vol.10Issue(10)：967-974,8.DOI:10.12379/j.issn.2096-1057.2024.10.11
基于小批量随机梯度下降法的SVM训练隐私保护方案
Privacy-preserving Scheme for SVM Training Based on Mini-batch SGD
摘要
Abstract
When using a support vector machine(SVM)to process sensitive data,privacy protection is very important.The existing SVM privacy-preserving schemes are trained based on batch gradient descent(BGD)algorithm,and they have huge computational overhead.To solve this problem,this paper proposed a privacy-preserving scheme for SVM training based on mini-batch stochastic gradient descent(Mini-batch SGD).Firstly,it designed the SVM training algorithm based on Mini-batch SGD.Then,on this basis,it perturbed the model weights by multiplication,used the hardness assumption of integer factorization to ensure the privacy of the model,engaged the homomorphic cryptosystem to encrypt the data,performed SVM training,and then applied the homomorphic hash function for verification.Finally,it constructed the SVM privacy-preserving scheme.Against security threats,the paper demonstrated data privacy,model privacy,and model correctness.It carried out simulation experiments and analysis of the scheme.The results show that the proposed scheme can save 92.4％of the computation time on average,while the classification performance is close to the existing schemes.关键词
小批量随机梯度下降法/支持向量机/同态加密/同态哈希函数/隐私保护Key words
Mini-batch SGD/SVM/homomorphic encryption/homomorphic hash function/privacy-preserving分类
信息技术与安全科学引用本文复制引用
王杰昌,刘玉岭,张平,刘牧华,赵新辉..基于小批量随机梯度下降法的SVM训练隐私保护方案[J].信息安全研究,2024,10(10):967-974,8.基金项目
国家自然科学基金项目(62102134) （62102134）
基础加强计划技术领域基金项目(2021-JCJQ-JJ-0908) （2021-JCJQ-JJ-0908）
河南省科技攻关项目(232102210138,232102210130,232102320309) （232102210138,232102210130,232102320309）
龙门实验室重大科技项目(231100220300) （231100220300）
河南省高等学校重点科研项目(23A520046,23A413005) （23A520046,23A413005）
