研究隐私参数对图像分类模型精度的影响，找到高效平衡隐私与性能的设置。
Impact of Privacy Parameters on Deep Learning Models for Image Classification
- 在CIFAR-10上测试五种模型，用差分隐私机制控制隐私损失
- EfficientNet在ε=5.0时达59.63%准确率，表现最佳
- 适合关注隐私保护下模型实用性的研究者参考
本项目旨在构建针对CIFAR-10数据集的差分私密深度学习图像分类模型，并分析不同隐私参数对模型准确率的影响。我们实现了五种深度学习模型（ConvNet、ResNet18、EfficientNet、ViT、DenseNet121）和三种监督分类器（K-近邻、朴素贝叶斯、支持向量机），在多种配置下评估其性能。目前表现最优的模型为EfficientNet，其在使用Adam优化器、批量大小256、训练轮数100、ε值5.0、学习率1e-3、裁剪阈值1.0及噪声倍数0.912的条件下，测试准确率达到59.63%。
原文摘要 · Abstract (English)
The project aims to develop differentially private deep learning models for image classification on CIFAR-10 datasets \cite{cifar10} and analyze the impact of various privacy parameters on model accuracy. We have implemented five different deep learning models, namely ConvNet, ResNet18, EfficientNet, ViT, and DenseNet121 and three supervised classifiers namely K-Nearest Neighbors, Naive Bayes Classifier and Support Vector Machine. We evaluated the performance of these models under varying settings. Our best performing model to date is EfficientNet with test accuracy of $59.63\%$ with the following parameters (Adam optimizer, batch size 256, epoch size 100, epsilon value 5.0, learning rate $1e-3$, clipping threshold 1.0, and noise multiplier 0.912).
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。