9,243 Hits in 5.1 sec

Optimal Accuracy-Privacy Trade-Off for Secure Multi-Party Computations [article]

Patrick Ah-Fat, Michael Huth
2018 arXiv   pre-print
Finally, we investigate the inherent compromise between accuracy of computation and privacy of inputs and we demonstrate how to realise such optimal trade-offs.  ...  The purpose of Secure Multi-Party Computation is to enable protocol participants to compute a public function of their private inputs while keeping their inputs secret, without resorting to any trusted  ...  This presents an inherent trade-off between the accuracy of the output and the privacy of the inputs. We will understand this trade-off in detail in Section 8.  ... 
arXiv:1803.00436v1 fatcat:rofor4av2veihcv2lt75jpzpbq

Achieving Secure and Differentially Private Computations in Multiparty Settings [article]

Abbas Acar, Z. Berkay Celik, Hidayet Aksu, A. Selcuk Uluagac, Patrick McDaniel
2017 arXiv   pre-print
Secure multiparty computation (SMC) is a viable panacea for this, allowing distributed parties to make computations while the parties learn nothing about their data, but the final result.  ...  The protocol implements homomorphic encryption for SMC and functional mechanism for DP to achieve the desired security and privacy guarantees.  ...  Accuracy Analysis We evaluate the accuracy-privacy trade-off of distributed evaluation of differential privacy on linear regression. Specif- Figure 6 .  ... 
arXiv:1707.01871v1 fatcat:bcwxjj2yg5amdpnhd225t6igke

Privacy-Preserving Federated Neural Network Learning for Disease-Associated Cell Classification [article]

Sinem Sav, Jean-Philippe Bossuat, Juan R. Troncoso-Pastoriza, Manfred Claassen, Jean-Pierre Hubaux
2022 bioRxiv   pre-print
Our work guarantees patient privacy and ensures data utility for efficient multi-center studies involving complex healthcare data.  ...  Our solution achieves an accuracy comparable to the one obtained with the centralized solution, with an improvement of at least one-order-of-magnitude in execution time with respect to prior secure solutions  ...  We also thank Shufan Wang and Joao Sa Sousa for their contribution on the experiments and benchmarking.  ... 
doi:10.1101/2022.01.10.475610 fatcat:uokre4qusnfmbbzaqzrmb6in5e

DP-Cryptography: Marrying Differential Privacy and Cryptography in Emerging Applications [article]

Sameer Wagh, Xi He, Ashwin Machanavajjhala, Prateek Mittal
2020 arXiv   pre-print
First, we survey solutions that combine cryptographic primitives like secure computation and anonymous communication with differential privacy to give alternatives to the LDP model that avoid a trusted  ...  server as in SDP but close the gap in accuracy.  ...  On the other hand, systems such as RAPPOR allow for approximate computation of statistics and primarily provide a privacy-utility trade-off.  ... 
arXiv:2004.08887v1 fatcat:xkwpazhrd5gyndl2sbs3pquxni

Differentially Private Secure Multi-Party Computation for Federated Learning in Financial Applications [article]

David Byrd, Antigoni Polychroniadou
2020 arXiv   pre-print
This could produce more accurate models for important tasks like optimal trade execution, credit origination, or fraud detection.  ...  A system can further reduce the coordinating server's ability to recover private client information, without additional accuracy loss, by also including secure multiparty computation.  ...  Morgan makes no representation and warranty whatsoever and disclaims all liability, for the completeness, accuracy or reliability of the information contained herein.  ... 
arXiv:2010.05867v1 fatcat:gtrmrbyeqbcltmcq2viruifezy

Secure Multi-party Computation of Differentially Private Median

Jonas Böhler, Florian Kerschbaum
2020 USENIX Security Symposium  
data universe (secure multi-party computation).  ...  ) and convex optimizations for machine learning.  ...  Preliminaries In the following, we introduce preliminaries for differential privacy and secure multi-party computation.  ... 
dblp:conf/uss/BohlerK20 fatcat:iesp7lhvyjabvjqsiygjrw6z2m

Mining multiple private databases using a kNN classifier

Li Xiong, Subramanyam Chitti, Ling Liu
2007 Proceedings of the 2007 ACM symposium on Applied computing - SAC '07  
A salient feature of our approach is that it offers a trade-off between accuracy, efficiency and privacy through multi-round protocols.  ...  Modern electronic communication has collapsed geographical boundaries for global information sharing but often at the expense of data security and privacy boundaries.  ...  At the other end, we have the secure multi-party computation protocols, using which we can construct classifiers which are provably secure in the sense that they reveal the least amount of information  ... 
doi:10.1145/1244002.1244102 dblp:conf/sac/XiongCL07 fatcat:th4esaqpf5b3lcozc5dlegrfwi

Distributed Learning without Distress: Privacy-Preserving Empirical Risk Minimization

Bargav Jayaraman, Lingxiao Wang, David Evans, Quanquan Gu
2018 Neural Information Processing Systems  
For both methods, we show that the noise can be reduced in the multi-party setting by adding the noise inside the secure computation after aggregation, asymptotically improving upon the best previous results  ...  In our output perturbation method, the parties combine local models within a secure computation and then add the required differential privacy noise before revealing the model.  ...  Background on Differential Privacy and Multi-Party Computation This section introduces differential privacy (including the zero-concentrated differential privacy notion we use), and secure multi-party  ... 
dblp:conf/nips/JayaramanW0G18 fatcat:smqfiv7gvzdy7kstfbg6vl5wam

Differentially-Private Multi-Party Sketching for Large-Scale Statistics

Seung Geol Choi, Dana Dachman-soled, Mukul Kulkarni, Arkady Yerukhimovich
2020 Proceedings on Privacy Enhancing Technologies  
among efficiency, privacy and accuracy of our implementation for varying numbers of parties and input sizes.Finally, we generalize our approach for the LogLog sketch and obtain a general framework for  ...  constructing multi-party differentially private protocols for several other sketching algorithms.  ...  All DP mechanisms inherently suffer from a trade-off between accuracy and privacy.  ... 
doi:10.2478/popets-2020-0047 fatcat:tuadgjwozrfrtdcgmvocxmqsnq

Towards Privacy-Preserving Neural Architecture Search [article]

Fuyi Wang and Leo Yu Zhang and Lei Pan and Shengshan Hu and Robin Doss
2022 arXiv   pre-print
To address these issues, we propose a privacy-preserving neural architecture search (PP-NAS) framework based on secure multi-party computation to protect users' data and the model's parameters/hyper-parameters  ...  Extensive analyses and experiments demonstrate PP-NAS's superiority in security, efficiency, and accuracy.  ...  To protect user privacy, privacy-preserving machine learning (PPML) based on secure multi-party computation (MPC) [6] - [9] has been proposed and implemented.  ... 
arXiv:2204.10958v1 fatcat:ta7neeob4newdgridsdvcxz73i

SAQE: Practical Privacy-Preserving Approximate Query Processing for Data Federations

Johes Bater, Yongjoo Park, Xi He, Xiao Wang, Jennie Rogers
2020 Proceedings of the VLDB Endowment  
Experimentally, we show that this result enables SAQE to trade-off among these three criteria to scale its query processing to very large datasets with accuracy bounds dependent only on sample size, and  ...  , secure computation, and approximate query processing -in a novel and principled way.  ...  Finally, we examine the effectiveness of our cost model and optimization in relation to the trade-offs between accuracy, privacy, and efficiency.  ... 
dblp:journals/pvldb/BaterP0WR20 fatcat:sfg3d4wkg5h2jfpwfhz272m364

Enhancing the Privacy of Federated Learning with Sketching [article]

Zaoxing Liu, Tian Li, Virginia Smith, Vyas Sekar
2019 arXiv   pre-print
To better optimize these trade-offs, we propose that sketching algorithms have a unique advantage in that they can provide both privacy and performance benefits while maintaining accuracy.  ...  Our initial findings show that it is possible to provide strong privacy guarantees for federated learning without sacrificing performance or accuracy.  ...  For instance, [10] brings secure multi-party computation (SMC) into federated learning.  ... 
arXiv:1911.01812v1 fatcat:mquqgd2ykjepdidtf5bx4rkkpq

Modeling Privacy Aware Information Sharing Systems: A Formal and General Approach

Fabio Martinelli, Andrea Saracino, Mina Sheikhalishahi
2016 2016 IEEE Trustcom/BigDataSE/ISPA  
, and includes a novel method to compute the trade-off between privacy and accuracy.  ...  This work also proposes a methodology to choose, case-by-case, the privacy mechanism which maximizes the trade-off between privacy and accuracy.  ...  To preserve privacy, in this architecture information is shared and analysis is performed by mean of secure multi-party computation algorithms, mainly based on homomorphic encryption, which allows data  ... 
doi:10.1109/trustcom.2016.0137 dblp:conf/trustcom/MartinelliSS16 fatcat:o7avy2nxobbl5fuxtht3rad7nu

Privacy-Preserving Machine Learning: Methods, Challenges and Directions [article]

Runhua Xu, Nathalie Baracaldo, James Joshi
2021 arXiv   pre-print
Hence, well-designed privacy-preserving ML (PPML) solutions are critically needed for many emerging applications.  ...  Usually, a well-performing ML model relies on a large volume of training data and high-powered computational resources.  ...  -public key infrastructure; SS -secret sharing; 2PC -secure two-party computation; MPC -secure multi-party computation.  ... 
arXiv:2108.04417v2 fatcat:pmxmsbs2gvh6nd4jadcz4dnsrq

Privacy Preserving Image Registration [article]

Riccardo Taiello, Melek Önen, Olivier Humbert, Marco Lorenzi
2022 arXiv   pre-print
We derive our privacy preserving image registration framework by extending classical registration paradigms to account for advanced cryptographic tools, such as secure multi-party computation and homomorphic  ...  We demonstrate our privacy preserving framework in linear and non-linear registration problems, evaluating its accuracy and scalability with respect to standard image registration.  ...  Secure Computation Secure Multi-Party Computation.  ... 
arXiv:2205.10120v4 fatcat:naalcnlbqrb3lj3wbnrl2qr52u
« Previous Showing results 1 — 15 out of 9,243 results