High-Precision Privacy-Preserving Real-Valued Function Evaluation
Autor: | Stanislav Peceny, Christina Boura, Ilaria Chillotti, Nicolas Gama, Alexander Petric, Dimitar Jetchev |
---|---|
Přispěvatelé: | Laboratoire de Mathématiques de Versailles (LMV), Université de Versailles Saint-Quentin-en-Yvelines (UVSQ)-Université Paris-Saclay-Centre National de la Recherche Scientifique (CNRS) |
Jazyk: | angličtina |
Rok vydání: | 2018 |
Předmět: |
050101 languages & linguistics
Floating point Computer science Computation 05 social sciences 02 engineering and technology Sigmoid function Logistic regression Real-valued function 0202 electrical engineering electronic engineering information engineering Rare events 020201 artificial intelligence & image processing 0501 psychology and cognitive sciences [MATH]Mathematics [math] Classifier (UML) Fourier series Algorithm |
Zdroj: | In Proceedings of Financial Cryptography and Data Security-(FC) 2018 In Proceedings of Financial Cryptography and Data Security-(FC) 2018, 2018, Nieuwpoort, Curaçao. ⟨10.1007/978-3-662-58387-6_10⟩ Financial Cryptography and Data Security ISBN: 9783662583869 Financial Cryptography |
DOI: | 10.1007/978-3-662-58387-6_10⟩ |
Popis: | We propose a novel multi-party computation protocol for evaluating continuous real-valued functions with high numerical precision. Our method is based on approximations with Fourier series and uses at most two rounds of communication during the online phase. For the offline phase, we propose a trusted-dealer and honest-but-curious aided solution, respectively. We apply our algorithm to train a logistic regression classifier via a variant of Newton’s method (known as IRLS) to compute unbalanced classification problems that detect rare events and cannot be solved using previously proposed privacy-preserving optimization algorithms (e.g., based on piecewise-linear approximations of the sigmoid function). Our protocol is efficient as it can be implemented using standard quadruple-precision floating point arithmetic. We report multiple experiments and provide a demo application that implements our algorithm for training a logistic regression model. |
Databáze: | OpenAIRE |
Externí odkaz: |