Search for: All records

Most existing approaches of differentially private (DP) machine learning focus on private training. Despite its many advantages, private training lacks the flexibility in adapting to incremental changes to the training dataset such as deletion requests from exercising GDPR’s right to be forgotten. We revisit a long-forgotten alternative, known as private prediction, and propose a new algorithm named Individual Kernelized Nearest Neighbor (Ind-KNN). Ind-KNN is easily updatable over dataset changes and it allows precise control of the Rényi DP at an individual user level — a user’s privacy loss is measured by the exact amount of her contribution to predictions; and a user is removed if her prescribed privacy budget runs out. Our results show that Ind-KNN consistently improves the accuracy over existing private prediction methods for a wide range of epsilon on four vision and language tasks. We also illustrate several cases under which Ind-KNN is preferable over private training with NoisySGD. 

Neural network robustness has become a central topic in machine learning in recent years. Most training algorithms that improve the model’s robustness to adversarial and common corruptions also introduce a large computational overhead, requiring as many as ten times the number of forward and backward passes in order to converge. To combat this inefficiency, we propose BulletTrain — a boundary example mining technique to drastically reduce the computational cost of robust training. Our key observation is that only a small fraction of examples are beneficial for improving robustness. BulletTrain dynamically predicts these important examples and optimizes robust training algorithms to focus on the important examples. We apply our technique to several existing robust training algorithms and achieve a 2.2× speed-up for TRADES and MART on CIFAR-10 and a 1.7× speed-up for AugMix on CIFAR-10-C and CIFAR-100-C without any reduction in clean and robust accuracy. 

We propose an intriguingly simple method for the construction of adversarial images in the black-box setting. In constrast to the white-box scenario, constructing black-box adversarial im- ages has the additional constraint on query bud- get, and efficient attacks remain an open prob- lem to date. With only the mild assumption of continuous-valued confidence scores, our highly query-efficient algorithm utilizes the following simple iterative principle: we randomly sample a vector from a predefined orthonormal basis and either add or subtract it to the target image. De- spite its simplicity, the proposed method can be used for both untargeted and targeted attacks – resulting in previously unprecedented query effi- ciency in both settings. We demonstrate the effi- cacy and efficiency of our algorithm on several real world settings including the Google Cloud Vision API. We argue that our proposed algorithm should serve as a strong baseline for future black- box attacks, in particular because it is extremely fast and its implementation requires less than 20 lines of PyTorch code.