skip to main content

This content will become publicly available on September 1, 2024

Title: Deep Neural Network Quantization Framework for Effective Defense against Membership Inference Attacks
Machine learning deployment on edge devices has faced challenges such as computational costs and privacy issues. Membership inference attack (MIA) refers to the attack where the adversary aims to infer whether a data sample belongs to the training set. In other words, user data privacy might be compromised by MIA from a well-trained model. Therefore, it is vital to have defense mechanisms in place to protect training data, especially in privacy-sensitive applications such as healthcare. This paper exploits the implications of quantization on privacy leakage and proposes a novel quantization method that enhances the resistance of a neural network against MIA. Recent studies have shown that model quantization leads to resistance against membership inference attacks. Existing quantization approaches primarily prioritize performance and energy efficiency; we propose a quantization framework with the main objective of boosting the resistance against membership inference attacks. Unlike conventional quantization methods whose primary objectives are compression or increased speed, our proposed quantization aims to provide defense against MIA. We evaluate the effectiveness of our methods on various popular benchmark datasets and model architectures. All popular evaluation metrics, including precision, recall, and F1-score, show improvement when compared to the full bitwidth model. For example, for ResNet on Cifar10, our experimental results show that our algorithm can reduce the attack accuracy of MIA by 14%, the true positive rate by 37%, and F1-score of members by 39% compared to the full bitwidth network. Here, reduction in true positive rate means the attacker will not be able to identify the training dataset members, which is the main goal of the MIA.  more » « less
Award ID(s):
2413046 2047384
Author(s) / Creator(s):
Publisher / Repository:
Date Published:
Journal Name:
Page Range / eLocation ID:
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. In the field of multi-agent autonomous transportation, such as automated payload delivery or highway on-ramp merging, agents routinely exchange knowledge to optimize their shared objective and adapt to environmental novelties through Cooperative Multi-Agent Reinforcement Learning (CMARL) algorithms. This knowledge exchange between agents allows these systems to operate efficiently and adapt to dynamic environments. However, this cooperative learning process is susceptible to adversarial poisoning attacks, as highlighted by contemporary research. Particularly, the poisoning attacks where malicious agents inject deceptive information camouflaged within the differential noise, a pivotal element for differential privacy (DP)-based CMARL algorithms, pose formidable challenges to identify and overcome. The consequences of not addressing this issue are far-reaching, potentially jeopardizing safety-critical operations and the integrity of data privacy in these applications. Existing research has strived to develop anomaly detection-based defense models to counteract conventional poisoning methods. Nonetheless, the recurring necessity for model offloading and retraining with labeled anomalous data undermines their practicality, considering the inherently dynamic nature of the safety-critical autonomous transportation applications. Further, it is imperative to maintain data privacy, ensure high performance, and adapt to environmental changes. Motivated by these challenges, this paper introduces a novel defense mechanism against stealthy adversarial poisoning attacks in the autonomous transportation domain, termed Reinforcing Autonomous Multi-agent Protection through Adversarial Resistance in Transportation (RAMPART). Leveraging a GAN model at each local node, RAMPART effectively filters out malicious advice in an unsupervised manner, whilst generating synthetic samples for each state-action pair to accommodate environmental uncertainties and eliminate the need for labeled training data. Our extensive experimental analysis, conducted in a Private Payload Delivery Network (PPDN) —a common application in the autonomous multi-agent transportation domain—demonstrates thatRAMPART successfully defends against a DP-exploited poisoning attack with a\(30\% \)attack ratio, achieving an F1 score of 0.852 and accuracy of\(96.3\% \)in heavy-traffic environments.

    more » « less
  2. We study the membership inference (MI) attack against classifiers, where the attacker's goal is to determine whether a data instance was used for training the classifier. Through systematic cataloging of existing MI attacks and extensive experimental evaluations of them, we find that a model's vulnerability to MI attacks is tightly related to the generalization gap -- the difference between training accuracy and test accuracy. We then propose a defense against MI attacks that aims to close the gap by intentionally reduces the training accuracy. More specifically, the training process attempts to match the training and validation accuracies, by means of a new {\em set regularizer} using the Maximum Mean Discrepancy between the softmax output empirical distributions of the training and validation sets. Our experimental results show that combining this approach with another simple defense (mix-up training) significantly improves state-of-the-art defense against MI attacks, with minimal impact on testing accuracy. 
    more » « less
  3. Explainability is increasingly recognized as an enabling technology for the broader adoption of machine learning (ML), particularly for safety-critical applications. This has given rise to explainable ML, which seeks to enhance the explainability of neural networks through the use of explanators. Yet, the pursuit for better explainability inadvertently leads to increased security and privacy risks. While there has been considerable research into the security risks of explainable ML, its potential privacy risks remain under-explored. To bridge this gap, we present a systematic study of privacy risks in explainable ML through the lens of membership inference. Building on the observation that, besides the accuracy of the model, robustness also exhibits observable differences among member samples and non-member samples, we develop a new membership inference attack. This attack extracts additional membership features from changes in model confidence under different levels of perturbations guided by the importance highlighted by the attribution maps in the explanators. Intuitively, perturbing important features generally results in a bigger loss in confidence for member samples. Using the member-non-member differences in both model performance and robustness, an attack model is trained to distinguish the membership. We evaluated our approach with seven popular explanators across various benchmark models and datasets. Our attack demonstrates there is non-trivial privacy leakage in current explainable ML methods. Furthermore, such leakage issue persists even if the attacker lacks the knowledge of training datasets or target model architectures. Lastly, we also found existing model and output-based defense mechanisms are not effective in mitigating this new attack. 
    more » « less
  4. Federated Learning (FL) allows individual clients to train a global model by aggregating local model updates each round. This results in collaborative model training while main-taining the privacy of clients' sensitive data. However, malicious clients can join the training process and train with poisoned data or send artificial model updates in targeted poisoning attacks. Many defenses to targeted poisoning attacks rely on anomaly-detection based metrics which remove participants that deviate from the majority. Similarly, aggregation-based defenses aim to reduce the impact of outliers, while L2-norm clipping tries to scale down the impact of malicious models. However, oftentimes these defenses misidentify benign clients as malicious or only work under specific attack conditions. In our paper, we examine the effectiveness of two anomaly -detection metrics on three different aggregation methods, in addition to the presence of L2-norm clipping and weight selection, across two different types of attacks. We also combine different defenses in order to examine their interaction and examine each defense when no attack is present. We found minimum aggregation to be the most effective defense against label-flipping attacks, whereas both minimum aggregation and geometric median worked well against distributed backdoor attacks. Using random weight selection significantly deteriorated defenses against both attacks, whereas the use of clipping made little difference. Finally, the main task accuracy was directly correlated with the BA in the label-flipping attack and generally was close to the MA in benign scenarios. However, in the DBA the MA and BA are inversely correlated and the MA fluctuates greatly. 
    more » « less
  5. The large model size, high computational operations, and vulnerability against membership inference attack (MIA) have impeded deep learning or deep neural networks (DNNs) popularity, especially on mobile devices. To address the challenge, we envision that the weight pruning technique will help DNNs against MIA while reducing model storage and computational operation. In this work, we propose a pruning algorithm, and we show that the proposed algorithm can find a subnetwork that can prevent privacy leakage from MIA and achieves competitive accuracy with the original DNNs. We also verify our theoretical insights with experiments. Our experimental results illustrate that the attack accuracy using model compression is up to 13.6% and 10% lower than that of the baseline and Min-Max game, accordingly.

    more » « less