{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,28]],"date-time":"2026-06-28T05:46:28Z","timestamp":1782625588231,"version":"3.54.5"},"reference-count":38,"publisher":"SAGE Publications","issue":"5","license":[{"start":{"date-parts":[[2020,2,29]],"date-time":"2020-02-29T00:00:00Z","timestamp":1582934400000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/journals.sagepub.com\/page\/policies\/text-and-data-mining-license"}],"content-domain":{"domain":["journals.sagepub.com"],"crossmark-restriction":true},"short-container-title":["Journal of Intelligent &amp; Fuzzy Systems"],"published-print":{"date-parts":[[2020,5,29]]},"abstract":"<jats:p>\n                    \u00a0Applications using Artificial Intelligence techniques demand a thorough assessment of different aspects of trust, namely, data and model privacy, reliability, robustness against adversarial attacks, fairness, and interpretability. While each of these aspects has been extensively studied in isolation, an understanding of the trade-offs between different aspects of trust is lacking. In this work, the trade-off between fault tolerance, privacy, and adversarial robustness is evaluated for Deep Neural Networks, by considering two adversarial settings under security and a privacy threat model. Specifically, this work studies the impact of training the model with input noise (Adversarial Robustness) and gradient noise (Differential Privacy) on Neural Network\u2019s fault tolerance. While adding noise to inputs, gradients or weights enhances fault tolerance, it is observed that adversarial robustness lowers fault tolerance due to increased overfitting. On the other hand, (\n                    <jats:italic>\u03b5<\/jats:italic>\n                    <jats:sub>\n                      <jats:italic>dp<\/jats:italic>\n                    <\/jats:sub>\n                    ,\n                    <jats:italic>\u03b4<\/jats:italic>\n                    <jats:sub>\n                      <jats:italic>dp<\/jats:italic>\n                    <\/jats:sub>\n                    )-Differentially Private models enhance the fault tolerance, measured using generalisation error, which theoretically has an upper bound of\n                    <jats:italic>e<\/jats:italic>\n                    <jats:sup>\n                      <jats:italic>\u03b5<\/jats:italic>\n                      <jats:sub>\n                        <jats:italic>dp<\/jats:italic>\n                      <\/jats:sub>\n                    <\/jats:sup>\n                    \u00a0-\u00a01\u00a0+\u00a0\n                    <jats:italic>\u03b4<\/jats:italic>\n                    <jats:sub>\n                      <jats:italic>dp<\/jats:italic>\n                    <\/jats:sub>\n                    . This novel study of the trade-offs between different aspects of trust is pivotal for training trustworthy Machine Learning models.\n                  <\/jats:p>","DOI":"10.3233\/jifs-179677","type":"journal-article","created":{"date-parts":[[2020,3,3]],"date-time":"2020-03-03T09:55:00Z","timestamp":1583229300000},"page":"5897-5907","update-policy":"https:\/\/doi.org\/10.1177\/sage-journals-update-policy","source":"Crossref","is-referenced-by-count":10,"title":["Fault tolerance of neural networks in adversarial settings"],"prefix":"10.1177","volume":"38","author":[{"given":"Vasisht","family":"Duddu","sequence":"first","affiliation":[{"name":"Indraprastha Institute of Information Technology, Delhi, India"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"N.","family":"Rajesh Pillai","sequence":"additional","affiliation":[{"name":"Scientific Analysis Group, Delhi, India"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"D. Vijay","family":"Rao","sequence":"additional","affiliation":[{"name":"Institute for Systems Studies and Analyses, Delhi, India"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Valentina E.","family":"Balas","sequence":"additional","affiliation":[{"name":"Aurel Vlaicu University of Arad, Arad, Romania"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"179","published-online":{"date-parts":[[2020,2,29]]},"reference":[{"key":"e_1_3_2_2_2","unstructured":"ShokriR. StrobelM. and ZickY. Privacy risks of explaining machine learning models arXiv preprint arXiv:1907.00164 2019."},{"key":"e_1_3_2_3_2","doi-asserted-by":"crossref","unstructured":"SongL. ShokriR. and MittalP. Privacy risks of securing machine learning models against adversarial examples arXiv preprint arXiv:1905.10291 2019.","DOI":"10.1145\/3319535.3354211"},{"key":"e_1_3_2_4_2","unstructured":"JagielskiM. KearnsM. MaoJ. OpreaA. RothA. MalvajerdiS.S. and UllmanJ. Differentially private fair learning. In Kamalika Chaudhuri and Ruslan Salakhutdinov editors Proceedings of the 36th International Conference on Machine Learning volume 97 of Proceedings of Machine Learning Research pages 3000\u20133008 Long Beach California USA 09\u201315 Jun 2019. PMLR."},{"key":"e_1_3_2_5_2","unstructured":"EtmannC. LunzS. MaassP. and SchoenliebC. On the connection between adversarial robustness and saliency map interpretability. In Kamalika Chaudhuri and R. Salakhutdinov editors Proceedings of the 36th International Conference on Machine Learning volume 97 of Proceedings of Machine Learning Research pages 1823\u20131832 Long Beach California USA 09\u201315 Jun 2019. PMLR."},{"key":"e_1_3_2_6_2","unstructured":"DudduV. RaoD.V. and BalasV.E. Adversarial fault tolerant training for deep neural networks arXiv preprint arXiv:1907.03103 2019."},{"issue":"8","key":"e_1_3_2_7_2","first-page":"2018","article-title":"Regularizing multilayer perceptron for robustness,\u2013Aug","volume":"48","author":"Dey P.","unstructured":"DeyP., NagK., PalT. and Pal.N.R., Regularizing multilayer perceptron for robustness,\u2013Aug, IEEE Transactions on Systems, Man, and Cybernetics: Systems48(8) (1266), 2018.","journal-title":"IEEE Transactions on Systems, Man, and Cybernetics: Systems"},{"key":"e_1_3_2_8_2","doi-asserted-by":"publisher","DOI":"10.1109\/72.712162"},{"key":"e_1_3_2_9_2","doi-asserted-by":"publisher","DOI":"10.1109\/72.105414"},{"key":"e_1_3_2_10_2","doi-asserted-by":"publisher","DOI":"10.1162\/neco.1995.7.1.108"},{"key":"e_1_3_2_11_2","doi-asserted-by":"crossref","unstructured":"GanjuK. WangQ. YangW. GunterC.A. and BorisovN. Property inference attacks on fully connected neural networks using permutation invariant representations. In Proceedings of the 2018 ACM SIGSAC Conference on Computer and Communications Security CCS \u201918 pages 619\u2013633 New York NY USA 2018. ACM.","DOI":"10.1145\/3243734.3243834"},{"key":"e_1_3_2_12_2","doi-asserted-by":"crossref","unstructured":"ShokriR. StronatiM. SongC. and ShmatikovV. Membership inference attacks against machine learning models. In 2017 IEEE Symposium on Security and Privacy (SP) pages 3\u201318 May 2017.","DOI":"10.1109\/SP.2017.41"},{"key":"e_1_3_2_13_2","doi-asserted-by":"crossref","unstructured":"SalemA. ZhangY. HumbertM. FritzM. and BackesM. Ml-leaks: Model and data independent membership inference attacks and defenses on machine learning models. In Annual Network and Distributed System Security Symposium (NDSS) 2019. to appear.","DOI":"10.14722\/ndss.2019.23119"},{"key":"e_1_3_2_14_2","doi-asserted-by":"crossref","unstructured":"DworkC. and RothA. The algorithmic foundations of differential privacy 2014.","DOI":"10.1561\/9781601988195"},{"key":"e_1_3_2_15_2","unstructured":"AbadiM. ChuA. GoodfellowI. McMahanH.B. MironovI. TalwarK. and ZhangL. Deep learning with differential privacy. In Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security CCS\u201916 pages 308\u2013318 NewYork NY USA 2016.ACM."},{"key":"e_1_3_2_16_2","doi-asserted-by":"publisher","DOI":"10.1162\/089976600300014782"},{"key":"e_1_3_2_17_2","unstructured":"WangJ. ChangQ. ChangQ. LiuY. and PalN.R. Weight noise injection-based mlps with group lasso penalty: Asymptotic convergence and application to node pruning IEEE Transactions on Cybernetics pages 1\u201319 2018."},{"key":"e_1_3_2_18_2","doi-asserted-by":"publisher","DOI":"10.1109\/72.363479"},{"key":"e_1_3_2_19_2","doi-asserted-by":"crossref","unstructured":"BeiuV. RohatinoviciN.C. D\u0103u\u015fL. and BalasV.E. Transport reliability on axonal cytoskeleton. In 2017 14th International Conference on Engineering of Modern Electric Systems (EMES) pages 160\u2013163 June 2017.","DOI":"10.1109\/EMES.2017.7980404"},{"key":"e_1_3_2_20_2","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2018.2828036"},{"key":"e_1_3_2_21_2","doi-asserted-by":"publisher","DOI":"10.3233\/IFS-120762"},{"key":"e_1_3_2_22_2","doi-asserted-by":"publisher","DOI":"10.14429\/dsj.68.12371"},{"key":"e_1_3_2_23_2","unstructured":"MadryA. MakelovA. SchmidtL. TsiprasD. and VladuA. Towards deep learning models resistant to adversarial attacks. In International Conference on Learning Representations 2018."},{"key":"e_1_3_2_24_2","unstructured":"ZhangH. YuY. JiaoJ. XingE. GhaouiL.E. and JordanM. Theoretically principled trade-off between robustness and accuracy. In Kamalika Chaudhuri and Ruslan Salakhutdinov editors Proceedings of the 36th International Conference on Machine Learning volume 97 of Proceedings of Machine Learning Research pages 7472\u20137482 Long Beach California USA 09\u201315 Jun 2019. PMLR."},{"key":"e_1_3_2_25_2","unstructured":"SinhaA. NamkoongH. and DuchiJ. Certifiable distributional robustness with principled adversarial training. In International Conference on Learning Representations 2018."},{"key":"e_1_3_2_26_2","unstructured":"MirmanM. GehrT. and VechevM. Differentiable abstract interpretation for provably robust neural networks. In Jennifer Dy and Andreas Krause editors Proceedings of the 35th International Conference on Machine Learning volume 80 of Proceedings of Machine Learning Research pages 3578\u20133586 Stockholm sm\u00e4ssan Stockholm Sweden 10\u201315 Jul 2018. PMLR."},{"key":"e_1_3_2_27_2","unstructured":"DworkC. Differential privacy: A survey of results. In Manindra Agrawal Dingzhu Du Zhenhua Duan and Angsheng Li editors Theory and Applications of Models of Computation pages 1\u201319 Berlin Heidelberg 2008. Springer Berlin Heidelberg."},{"key":"e_1_3_2_28_2","doi-asserted-by":"crossref","unstructured":"MironovI. R\u00e9nyi differential privacy. In 2017 IEEE 30th Computer Security Foundations Symposium (CSF) pages 263\u2013275 Aug 2017.","DOI":"10.1109\/CSF.2017.11"},{"key":"e_1_3_2_29_2","doi-asserted-by":"crossref","unstructured":"FredriksonM. JhaS. and RistenpartT. Model inversion attacks that exploit confidence information and basic countermeasures. In Proceedings of the 22Nd ACM SIGSAC Conference on Computer and Communications Security CCS \u201915 pages 1322\u20131333 New York NY USA 2015. ACM.","DOI":"10.1145\/2810103.2813677"},{"key":"e_1_3_2_30_2","unstructured":"SalemA. BhattacharyyaA. BackesM. FritzM. and ZhangY. Updates-leak: Data set inference and reconstruction attacks in online learning. CoRR abs\/1904.01067 2019."},{"key":"e_1_3_2_31_2","unstructured":"CarliniN. LiuC. Erlingsson\u00da. KosJ. and SongD. The secret sharer: Evaluating and testing unintended memorization in neural networks. In 28th USENIX Security Symposium (USENIX Security 19) pages 267\u2013284 Santa Clara CA August 2019. USENIX Association."},{"key":"e_1_3_2_32_2","unstructured":"JayaramanB. and EvansD. Evaluating differentially private machine learning in practice. In 28th USENIX Security Symposium(USENIX Security 19) pages 1895\u20131912 Santa Clara CA August 2019. USENIX Association."},{"key":"e_1_3_2_33_2","doi-asserted-by":"crossref","unstructured":"NasrM. ShokriR. and HoumansadrA. Machine learning with membership privacy using adversarial regularization. In Proceedings of the 2018 ACM SIGSAC Conference on Computer and Communications Security CCS \u201918 pages 634\u2013646 New York NY USA 2018. ACM.","DOI":"10.1145\/3243734.3243855"},{"key":"e_1_3_2_34_2","doi-asserted-by":"crossref","unstructured":"JiaJ. SalemA. BackesM. ZhangY. and GongN.Z. Memguard: Defending against blackbox membership inference attacks via adversarial examples. arXiv preprint arXiv:1909.10594 2019.","DOI":"10.1145\/3319535.3363201"},{"key":"e_1_3_2_35_2","unstructured":"PapernotN. SongS. MironovI. RaghunathanA. TalwarK. and ErlingssonU. Scalable private learning with PATE. In International Conference on Learning Representations 2018."},{"key":"e_1_3_2_36_2","doi-asserted-by":"publisher","DOI":"10.1109\/TNN.2010.2046179"},{"key":"e_1_3_2_37_2","doi-asserted-by":"publisher","DOI":"10.1109\/72.105415"},{"key":"e_1_3_2_38_2","doi-asserted-by":"publisher","DOI":"10.1109\/21.155944"},{"issue":"183","key":"e_1_3_2_39_2","first-page":"1","article-title":"Learning with differential privacy: Stability, learnability and the sufficiency and necessity of erm principle","volume":"17","author":"Wang Y.-X.","year":"2016","unstructured":"WangY.-X., LeiJ. and FienbergS.E., Learning with differential privacy: Stability, learnability and the sufficiency and necessity of erm principle, Journal of Machine Learning Research17(183) (2016), 1\u201340.","journal-title":"Journal of Machine Learning Research"}],"container-title":["Journal of Intelligent &amp; Fuzzy Systems"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/journals.sagepub.com\/doi\/pdf\/10.3233\/JIFS-179677","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/journals.sagepub.com\/doi\/full-xml\/10.3233\/JIFS-179677","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/journals.sagepub.com\/doi\/pdf\/10.3233\/JIFS-179677","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,29]],"date-time":"2026-04-29T09:41:15Z","timestamp":1777455675000},"score":1,"resource":{"primary":{"URL":"https:\/\/journals.sagepub.com\/doi\/10.3233\/JIFS-179677"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2020,2,29]]},"references-count":38,"journal-issue":{"issue":"5","published-print":{"date-parts":[[2020,5,29]]}},"alternative-id":["10.3233\/JIFS-179677"],"URL":"https:\/\/doi.org\/10.3233\/jifs-179677","relation":{},"ISSN":["1064-1246","1875-8967"],"issn-type":[{"value":"1064-1246","type":"print"},{"value":"1875-8967","type":"electronic"}],"subject":[],"published":{"date-parts":[[2020,2,29]]}}}