Secure Computation on Floating Point Numbers

By Mehrdad Aliasgari, Marina Blanton, Yihua Zhang, Aaron Steele. 2013

131 citations recorded by the source corpus.

How researchers see this paper

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

Citations in the corpus

Secure Evaluation of Quantized Neural Networks (2019, background): “Furthermore, the authors left as future work extending their techniques to the floating-point setting like those from [1], which would achieve more fidelity with respect to the evaluation of a model in the clear.”
Falcon: Honest-Majority Maliciously Secure Framework for Private Deep Learning (2020, methodology): “[40, 41] for choice of constants).”
Practical Covertly Secure MPC for Dishonest Majority - Or: Breaking the SPDZ Limits (2013, background): “Finally, the “shared bits” are useful in computing high level operation such as comparison, bit-decomposition, fixed and floating point operations as in [1, 8, 9].”
Efficient, Oblivious Data Structures for MPC (2014, methodology): “[1] provide a protocol that allows to compute the modulo operation of a secret number and two raised to the power of a secret number.”
Generalizing the SPDZ Compiler For Other Protocols (2018, methodology): “This library is based on a body of literature [2, 10, 11] that uses techniques such as statistical masking to implement such operations without having to rely solely on field-arithmetic circuits.”
Fast, Privacy Preserving Linear Regression over Distributed Datasets based on Pre-Distributed Data (2015, methodology): “The only problem for applying this technique is that the truncation protocol used before is probabilistic and so it would not allow propagation of the MACs, but this can be solved by using a deterministic truncation procedure [1] (at the cost of having to perform one execution of a comparison protoc”
Differentially private data aggregation with optimal utility (2014, methodology): “We show how to achieve provable DDP guarantees, while overcoming the previously discussed limitations, by leveraging recently proposed SMPC protocols for ﬂoating point numbers [8], ﬁxed point numbers [13], and integers [25].”
Combining Secret Sharing and Garbled Circuits for Efficient Private IEEE 754 Floating-Point Computations (2015, methodology): “Existing implementations based on secret sharing provide only near approximations to the IEEE 754 standard [1][20][27].”