We present two information leakage attacks that outperform previous work on membership inference against generative models. The first attack allows membership inference without assumptions on the type of the generative model. Contrary to previous evaluation metrics for generative models, like Kernel Density Estimation, it only considers samples of the model which are close to training data records. The second attack specifically targets Variational Autoencoders, achieving high membership inference accuracy. Furthermore, previous work mostly considers membership inference adversaries who perform single record membership inference. We argue for considering regulatory actors who perform set membership inference to identify the use of specific datasets for training. The attacks are evaluated on two generative model architectures, Generative Adversarial Networks (GANs) and Variational Autoen-coders (VAEs), trained on standard image datasets. Our results show that the two attacks yield success rates superior to previous work on most data sets while at the same time having only very mild assumptions. We envision the two attacks in combination with the membership inference attack type formalization as especially useful. For example, to enforce data privacy standards and automatically assessing model quality in machine learning as a service setups. In practice, our work motivates the use of GANs since they prove less vulnerable against information leakage attacks while producing detailed samples.

No abstract

Attacks that aim to identify the training data of neural networks represent a severe threat to the privacy of individuals in the training dataset. A possible protection is offered by anonymization of the training data or training function with differential privacy. However, data scientists can choose between local and central differential privacy, and need to select meaningful privacy parameters . A comparison of local and central differential privacy based on the privacy parameters furthermore potentially leads data scientists to incorrect conclusions, since the privacy parameters are reflecting different types of mechanisms. Instead, we empirically compare the relative privacy-accuracy trade-off of one central and two local differential privacy mechanisms under a whitebox membership inference attack. While membership inference only reflects a lower bound on inference risk and differential privacy formulates an upper bound, our experiments with several datasets show that the privacy-accuracy trade-off is similar for both types of mechanisms despite the large difference in their upper bound. This suggests that the upper bound is far from the practical susceptibility to membership inference. Thus, small in central differential privacy and large in local differential privacy result in similar membership inference risks, and local differential privacy can be a meaningful alternative to central differential privacy for differentially private deep learning besides the comparatively higher privacy parameters.

Hierarchical text classification consists of classifying text documents into a hierarchy of classes and sub-classes. Although Artificial Neural Networks have proved useful to perform this task, unfortunately, they can leak training data information to adversaries due to training data memorization. Using differential privacy during model training can mitigate leakage attacks against trained models, enabling the models to be shared safely at the cost of reduced model accuracy. This work investigates the privacy–utility trade-off in hierarchical text classification with differential privacy guarantees, and it identifies neural network architectures that offer superior trade-offs. To this end, we use a white-box membership inference attack to empirically assess the information leakage of three widely used neural network architectures. We show that large differential privacy parameters already suffice to completely mitigate membership inference attacks, thus resulting only in a moderate decrease in model utility. More specifically, for large datasets with long texts, we observed Transformer-based models to achieve an overall favorable privacy–utility trade-off, while for smaller datasets with shorter texts, convolutional neural networks are preferable.

There has been a large number of contributions on privacy-preserving smart metering with Differential Privacy, addressing questions from actual enforcement at the smart meter to billing at the energy provider. However, exploitation is mostly limited to application of cryptographic security means between smart meters and energy providers. We illustrate along the use case of privacy preserving load forecasting that Differential Privacy is indeed a valuable addition that unlocks novel information flows for optimization. We show that (i) there are large differences in utility along three selected forecasting methods, (ii) energy providers can enjoy good utility especially under the linear regression benchmark model, and (iii) households can participate in privacy preserving load forecasting with an individual re-identification risk < 60%, only 10% over random guessing. This is a pre-print of an article submitted to Springer Open Journal "Energy Informatics".