Membership Inference Attacks Against Machine Learning Models (2016-10-18T00:00:00.000000Z)

TL;DR

This work quantitatively investigates how machine learning models leak information about the individual data records on which they were trained and empirically evaluates the inference techniques on classification models trained by commercial "machine learning as a service" providers such as Google and Amazon.

Abstract

We quantitatively investigate how machine learning models leak information about the individual data records on which they were trained. We focus on the basic membership inference attack: given a data record and black-box access to a model, determine if the record was in the model's training dataset. To perform membership inference against a target model, we make adversarial use of machine learning and train our own inference model to recognize differences in the target model's predictions on the inputs that it trained on versus the inputs that it did not train on. We empirically evaluate our inference techniques on classification models trained by commercial "machine learning as a service" providers such as Google and Amazon. Using realistic datasets and classification tasks, including a hospital discharge dataset whose membership is sensitive from the privacy perspective, we show that these models can be vulnerable to membership inference attacks. We then investigate the factors that influence this leakage and evaluate mitigation strategies.

Authors

R. Shokri

4 papers

Vitaly Shmatikov

4 papers

M. Stronati

1 papers

TL;DR

Abstract

Authors

References40 items

Membership Privacy in MicroRNA-based Studies

Stealing Machine Learning Models via Prediction APIs

Deep Learning with Differential Privacy

Participatory Cultural Mapping Based on Collective Behavior Data in Location-Based Social Networks

Differentially Private Learning of Structured Discrete Distributions

Robust Traceability from Trace Amounts

Model Inversion Attacks that Exploit Confidence Information and Basic Countermeasures

Train faster, generalize better: Stability of stochastic gradient descent

Privacy-preserving deep learning

Privacy Preserving Data Mining

Distilling the Knowledge in a Neural Network

To Drop or Not to Drop: Robustness, Consistency and Differential Privacy Properties of Dropout

Differentially Private Bayesian Optimization

Crypto-Nets: Neural Networks over Encrypted Data

Private Empirical Risk Minimization: Efficient Algorithms and Tight Error Bounds

Privacy in Pharmacogenetics: An End-to-End Case Study of Personalized Warfarin Dosing

Private Predictive Analysis on Encrypted Medical Data

Hacking smart machines with smarter ones: How to extract meaningful data from machine learning classifiers

Privacy Aware Learning

Functional Mechanism: Regression Analysis under Differential Privacy

Privacy-Preserving ECG Classification With Branching Programs and Neural Networks

"You Might Also Like:" Privacy Risks of Collaborative Filtering

On the Difficulties of Disclosure Prevention in Statistical Databases or The Case for Differential Privacy

Learning in a Large Function Space: Privacy-Preserving Mechanisms for SVM Learning

Differentially Private Empirical Risk Minimization

Privacy-preserving logistic regression

Resolving Individuals Contributing Trace Amounts of DNA to Highly Complex Mixtures Using High-Density SNP Genotyping Microarrays

Privacy-preserving Naïve Bayes classification

Differential Privacy

Calibrating Noise to Sensitivity in Private Data Analysis

Privacy-preserving distributed k-means clustering over arbitrarily partitioned data

The Elements of Statistical Learning: Data Mining, Inference, and Prediction

Privacy preserving association rule mining in vertically partitioned data

Kernel Logistic Regression and the Import Vector Machine

Statistical inference considered harmful, " https://github.com/frankmcsherry/blog/blob/master/posts

Dropout: a simple way to prevent neural networks from overfitting

Learning Multiple Layers of Features from Tiny Images

Privacy-Preserving Multivariate Statistical Analysis: Linear Regression and Classification

Nature Genetics: doi:10.1038/ng.436Contents

How to define the type and architecture for shadow models? 2. How to obtain data for training shadow models?

Field of Study

Journal Information

Name

Page

Venue Information

Name

Type

URL

Alternate Names