Scalable Iterative Classification for Sanitizing Large-Scale Datasets

Li, Bo; Vorobeychik, Yevgeniy; Li, Muqun; Malin, Bradley

doi:10.1109/tkde.2016.2628180

Cited by 10 publications

(4 citation statements)

References 52 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Cheap ubiquitous computing enables the collection of massive amounts of personal data in a wide variety of domains [80]. In cloud platforms, most of the time, BD can not be reached because of the privacy and security concerns and effective tools are being deployed to detect and respond faster to cyberthreats, attacks, breaches of data.…”

Section: Discussionmentioning

confidence: 99%

Management of geo-distributed intelligence: Deep Insight as a Service (DINSaaS) on Forged Cloud Platforms (FCP)

Kuru

2021

Journal of Parallel and Distributed Computing

View full text Add to dashboard Cite

Section: Discussionmentioning

confidence: 99%

Management of geo-distributed intelligence: Deep Insight as a Service (DINSaaS) on Forged Cloud Platforms (FCP)

Kuru

2021

Journal of Parallel and Distributed Computing

View full text Add to dashboard Cite

“…Similarly, Jiang et al [38] have developed the t-plausibility algorithm to replace the known (labeled) sensitive identifiers within the documents and guarantee that the sanitized document is associated with at least t documents. Li et al [42] have proposed a game theoretic framework for automatic redacting sensitive information. In general, finding and redacting sensitive information with high accuracy is still challenging.…”

Section: Differentially Private Learning Methodsmentioning

confidence: 99%

Differentially Private Data Generative Models

Chen¹,

Xiang²,

Xue³

et al. 2018

Preprint

Self Cite

View full text Add to dashboard Cite

Deep neural networks (DNNs) have recently been widely adopted in various applications, and such success is largely due to a combination of algorithmic breakthroughs, computation resource improvements, and access to a large amount of data. However, the large-scale data collections required for deep learning often contain sensitive information, therefore raising many privacy concerns. Prior research has shown several successful attacks in inferring sensitive training data information, such as model inversion [24,26], membership inference [57], and generative adversarial networks (GAN) based leakage attacks against collaborative deep learning [36]. In this paper, to enable learning efficiency as well as to generate data with privacy guarantees and high utility, we propose a differentially private autoencoder-based generative model (DP-AuGM) and a differentially private variational autoencoder-based generative model (DP-VaeGM). We evaluate the robustness of two proposed models. We show that DP-AuGM can effectively defend against the model inversion, membership inference, and GAN-based attacks. We also show that DP-VaeGM is robust against the membership inference attack. We conjecture that the key to defend against the model inversion and GAN-based attacks is not due to differential privacy but the perturbation of training data. Finally, we demonstrate that both DP-AuGM and DP-VaeGM can be easily integrated with real-world machine learning applications, such as machine learning as a service and federated learning, which are otherwise threatened by the membership inference attack and the GAN-based attack, respectively.

show abstract

“…Despite its novel advantage in condensing the information of the entire dataset in a smaller dataset, dataset distillation is essentially a DNN model (see Section II). Previous studies [42], [47] have shown that DNN models (e.g., image classifiers, language models) are vulnerable to security and privacy attacks, such as adversarial attacks [21], [35], [55], inference attacks [18], [26], [54], [59], [61], [66], backdoor attacks [24], [60], [76], [86]. Yet, existing dataset distillation efforts [5], [52], [53], [89] mainly focus on designing new algorithms to distill a large dataset better.…”

Section: Introductionmentioning

confidence: 99%

Backdoor Attacks Against Dataset Distillation

Liu¹,

Li²,

Backes³

et al. 2023

Proceedings 2023 Network and Distributed System Security Symposium

View full text Add to dashboard Cite

Dataset distillation has emerged as a prominent technique to improve data efficiency when training machine learning models. It encapsulates the knowledge from a large dataset into a smaller synthetic dataset. A model trained on this smaller distilled dataset can attain comparable performance to a model trained on the original training dataset. However, the existing dataset distillation techniques mainly aim at achieving the best trade-off between resource usage efficiency and model utility. The security risks stemming from them have not been explored. This study performs the first backdoor attack against the models trained on the data distilled by dataset distillation models in the image domain. Concretely, we inject triggers into the synthetic data during the distillation procedure rather than during the model training stage, where all previous attacks are performed. We propose two types of backdoor attacks, namely NAIVEATTACK and DOORPING. NAIVEATTACK simply adds triggers to the raw data at the initial distillation phase, while DOORPING iteratively updates the triggers during the entire distillation procedure. We conduct extensive evaluations on multiple datasets, architectures, and dataset distillation techniques. Empirical evaluation shows that NAIVEATTACK achieves decent attack success rate (ASR) scores in some cases, while DOORPING reaches higher ASR scores (close to 1.0) in all cases. Furthermore, we conduct a comprehensive ablation study to analyze the factors that may affect the attack performance. Finally, we evaluate multiple defense mechanisms against our backdoor attacks and show that our attacks can practically circumvent these defense mechanisms. 1

show abstract

Scalable Iterative Classification for Sanitizing Large-Scale Datasets

Cited by 10 publications

References 52 publications

Management of geo-distributed intelligence: Deep Insight as a Service (DINSaaS) on Forged Cloud Platforms (FCP)

Management of geo-distributed intelligence: Deep Insight as a Service (DINSaaS) on Forged Cloud Platforms (FCP)

Differentially Private Data Generative Models

Backdoor Attacks Against Dataset Distillation

Contact Info

Product

Resources

About