Geodesic
Neural network obfuscation for computations over encrypted data
Prikladnaya Diskretnaya Matematika. Supplement, no. 13 (2020), pp. 85-93.

Voir la notice de l'article provenant de la source Math-Net.Ru

An approach to neural network cryptographic obfuscation of computations is proposed. Applying the previously obtained results on the property of strict obfuscation of indistinguishability for a neural network approximator, we propose to use neural networks to perform arithmetic and other operations on encrypted data, thus realizing the idea of using homomorphic encryption to perform trusted computations in an untrusted environment. The cryptographic properties of this mechanism are evaluated and compared with traditional approaches to encryption based on the secret key. The advantages and disadvantages of neural networks in relation to the problem of obfuscation and processing of encrypted data are discussed.
Keywords: artificial neural network, obfuscation, homomorphic encryption, secrecy estimation. 
@article{PDMA_2020_13_a24,
     author = {V. L. Eliseev},
     title = {Neural network obfuscation for computations over encrypted data},
     journal = {Prikladnaya Diskretnaya Matematika. Supplement},
     pages = {85--93},
     publisher = {mathdoc},
     number = {13},
     year = {2020},
     language = {ru},
     url = {http://geodesic.mathdoc.fr/item/PDMA_2020_13_a24/}
}
TY  - JOUR
AU  - V. L. Eliseev
TI  - Neural network obfuscation for computations over encrypted data
JO  - Prikladnaya Diskretnaya Matematika. Supplement
PY  - 2020
SP  - 85
EP  - 93
IS  - 13
PB  - mathdoc
UR  - http://geodesic.mathdoc.fr/item/PDMA_2020_13_a24/
LA  - ru
ID  - PDMA_2020_13_a24
ER  - 
%0 Journal Article
%A V. L. Eliseev
%T Neural network obfuscation for computations over encrypted data
%J Prikladnaya Diskretnaya Matematika. Supplement
%D 2020
%P 85-93
%N 13
%I mathdoc
%U http://geodesic.mathdoc.fr/item/PDMA_2020_13_a24/
%G ru
%F PDMA_2020_13_a24
V. L. Eliseev. Neural network obfuscation for computations over encrypted data. Prikladnaya Diskretnaya Matematika. Supplement, no. 13 (2020), pp. 85-93. http://geodesic.mathdoc.fr/item/PDMA_2020_13_a24/

[1] Nikolenko S., Kadurin A., Arkhangelskaya E., Glubokoe obuchenie. Pogruzhenie v mir neironnykh setei, Piter, SPb., 2020

[2] Alekseev D. V., “Priblizhenie funktsii neskolkikh peremennykh neironnymi setyami”, Fundamentalnaya i prikladnaya matematika, 15:3 (2009), 9–21

[3] Khaikin S., Neironnye seti: polnyi kurs, 2-e izd., Vilyams, M., 2008

[4] Focardi R., Luccio F. L., “Neural cryptanalysis of classical ciphers”, Proc. ICTCS, 2018 http://ceur-ws.org/Vol-2243/paper10.pdf

[5] Danziger M., Henriques M. A. A., “Improved cryptanalysis combining differential and artificial neural network schemes”, Intern. Telecommun. Symp. (ITS) (Sao Paulo, 2014), 1–5

[6] Turcanik M., “Using recurrent neural network for hash function generation”, Intern. Conf. Appl. Electronics (AE) (Pilsen, 2017), 1–4

[7] Lian S., Sun J., Wang Z., One-way hash function based on neural network, arXiv: 0707.4032

[8] Karras D. A., Zorkadis V., “On neural network techniques in the secure management of communication systems through improving and quality assessing pseudorandom stream generators”, Neural Networks, 16:5–6 (2003), 899–905 | DOI

[9] Kanter I., Kinzel W., Kanter E., “Secure exchange of information by synchronization of neural networks”, Europhys. Lett., 2002, no. 57, 141–147 | DOI

[10] Klimov A., Mityagin A., Shamir A., “Analysis of neural cryptography”, LNCS, 2502, 2002, 288–298 | MR

[11] Coutinho M., De Oliveira A. R., Borges F., et al., “Learning perfectly secure cryptography to protect communications with adversarial neural cryptography”, Sensors, 2018, no. 18, 1306 https://pubmed.ncbi.nlm.nih.gov/29695066/ | DOI

[12] Abadi M., Andersen D. G., Learning to protect communications with adversarial neural cryptography, arXiv: 1610.06918

[13] Xie P., Bilenko M., Finley T., et al., Crypto-nets: Neural networks over encrypted data, arXiv: 1412.6181

[14] Hecht-Nielsen R., “Kolmogorov's mapping neural network existence theorem”, IEEE First Annual Int. Conf. on Neural Networks (San Diego, 1987), v. 3, 11–13

[15] Dathathri R., Saarikivi O., Chen H., et al., “CHET: an optimizing compiler for fully-homomorphic neural-network inferencing”, Proc. PLDI 2019, ACM, N.Y., 2019, 142–156 | DOI

[16] Eliseev V. L., “Iskusstvennye neironnye seti kak mekhanizm obfuskatsii vychislenii”, Prikladnaya diskretnaya matematika. Prilozhenie, 2019, no. 12, 165–169

[17] Fergyuson N., Shnaier B., Prakticheskaya kriptografiya, Dialektika, M., 2005