Membership privacy: a unifying framework for privacy definitions

By Ninghui Li, Wahbeh H. Qardaji, Dong Su, Yi Wu, Weining Yang. 2013

180 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Sociology and Political Science

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Sociology and Political Science

Citations in the corpus

Stealing Machine Learning Models via Prediction APIs (2016, background): “Differential privacy (DP) [22] and its variants [34] have been explored as mechanisms for protecting, in particular, the privacy of ML training data [54].”
Knock Knock, Who's There? Membership Inference on Aggregate Location Data (2017, background): “[18] introduce a framework geared to formalize the notion of Positive vs Negative Membership Privacy, considering an adversary parameterized by her prior knowledge.”
Dependence Makes You Vulnberable: Differential Privacy Under Dependent Tuples (2016, background): “However, the privacy guarantees provided by the existing DP mechanisms are valid only under the assumption that the data tuples forming the database are pairwise independent (which is also implicitly assumed by the DP adversary model) [8], [21], [24], [25], [27], [40].”
Machine Learning with Membership Privacy using Adversarial Regularization (2018, background): “The relation between some different definitions of membership privacy and differential privacy is analyzed in the literature [32, 51].”
Privacy Risk in Machine Learning: Analyzing the Connection to Overfitting (2017, methodology): “We formalize quantitative advantage measures that capture the privacy risk to training data posed by two types of attack, namely membership inference [6], [7] and attribute inference [3], [4], [5], [8].”
Differential Privacy Techniques for Cyber Physical Systems: A Survey (2018, background): “Continuing this research line, researchers proposed membership privacy [28]”
When Relaxations Go Bad: "Differentially-Private" Machine Learning (2019, unknown): “Li et al. [42] point to this fact and provide a direct relationship between di ﬀ erential privacy and membership inference attacks.”
On Your Social Network De-anonymizablity: Quantification and Large Scale Evaluation with Seed Knowledge (2015, methodology): “Generally, the data anonymization techniques can be characterized into three classes: naive ID removal, k-anonymization (including randomly adding/deleting edges) [11][12], and differential privacy [13][14].”
Hiding in the Crowd: A Massively Distributed Algorithm for Private Averaging with Malicious Adversaries (2018, background): “This is related to the “no free lunch” theorem in data privacy (Kifer and Machanavajjhala, 2011), see also the discussions in Kasiviswanathan and Smith (2014); Li et al. (2013).”
The Unintended Consequences of Overfitting: Training Data Inference Attacks (2017, methodology): “This type of attack has been extensively studied in the adjacent area of genomics [19, 20], and more recently in the context of machine learning [6,7].”
DarkneTZ: towards model privacy at the edge using trusted execution environments (2020, background): “Nonetheless, recent work demonstrated that local models still leak private information [21, 33, 34, 47, 57, 61–63].”
Towards Measuring Membership Privacy (2017, background): “Most of these relaxations can be unified under the framework of membership privacy [24], which shows that protecting private information is equivalent to preventing an attacker from knowing whether an individual is included in the input dataset.”
Model Inversion Attacks that Exploit Confidence Information and Basic Countermeasures (2015, background): “[28] explore a privacy framework called membership privacy.”
Stolen Memories: Leveraging Model Memorization for Calibrated White-Box Membership Inference (2019, background): “This can be a privacy threat in itself, but vulnerability to MI has also come to be seen as a more general indicator of whether a model leaks private information [27, 38, 47], and is closely related to the guarantee provided by differential privacy [26].”
Membership Privacy in MicroRNA-based Studies (2016, background): “Definition 2 (Positive Membership Privacy [25]).”