Encrypted statistical machine learning: new privacy preserving methods

By Louis J. M. Aslett, Pedro M. Esperança, Chris C. Holmes. 2015

68 citations recorded by the source corpus.

We present two new statistical machine learning methods designed to learn on fully homomorphic encrypted (FHE) data. The introduction of FHE schemes following Gentry (2009) opens up the prospect of privacy preserving statistical machine learning analysis and modelling of encrypted data without compromising security constraints. We propose tailored algorithms for applying extremely random forests, involving a new cryptographic stochastic fraction estimator, and na\"{i}ve Bayes, involving a semi-parametric model for the class decision boundary, and show how they can be used to learn and predict from encrypted data. We demonstrate that these techniques perform competitively on a variety of classification data sets and provide detailed information about the computational practicalities of these and other FHE methods.

Subfields: Artificial Intelligence

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

Citations in the corpus

Privacy in Deep Learning: A Survey (2020, methodology): “There are only a handful of works that exploit solely homomorphic encryption for private training of machine learning models [98, 99, 25].”
CryptoDL: Deep Neural Networks over Encrypted Data (2017, methodology): “propose methods for implementing statistical machine learning over encrypted data and implement extremely random forests and Naive Bayes classifiers over 20 datasets [44].”
Privacy-preserving Machine Learning as a Service (2018, neutral): “While privacy-preserving machine learning based on secure multi-party computation techniques have been studied in the literature, those approaches focus on traditional machine learning algorithms such as linear regression [6], decision trees [2, 5, 53] and linear classifiers [5, 19].”
CryptoNets: Applying Neural Networks to Encrypted Data with High Throughput and Accuracy (2016, background): “Copyright 2016 by the author(s).”
A Survey on Security Threats and Defensive Techniques of Machine Learning: A Data Driven View (2018, background): “Regardless of a great advance of DP [101] and HE [49], [107]–[109], existing”
EPIC: Efficient Private Image Classification (or: Learning from the Masters) (2019, background): “[2], also work on homomorphically encrypted data.”
Preserving Multi-party Machine Learning with Homomorphic Encryption (2016, background): “Although implementing machine learning algorithms in encrypted domain has been studied in recent years (1; 2; 3; 5; 6; 11), there is not much work on multi-party machine learning in encrypted domain.”
Encrypted Accelerated Least Squares Regression (2017, methodology): “However, some recent work has also begun on developing new methodology inspired by traditional techniques, specifically tailored to homomorphic computation so that the whole analysis — model fitting and prediction — can be computed encrypted (Aslett et al., 2015b).”
Oblivious Neural Network Predictions via MiniONN Transformations (2017, methodology): “[4, 5] presented ways to train both simple models (e.”
Hiding in the Crowd: A Massively Distributed Algorithm for Private Averaging with Malicious Adversaries (2018, background): “However, they typically suﬀer from some drawbacks: (partially) homomorphic encryption schemes (Paillier, 1999; Graepel et al., 2012; Aslett et al., 2015) require the existence of a trusted third party, secure multi-party computation techniques (Yao, 1982; Lindell and Pinkas, 2009) are generally…”