{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,4]],"date-time":"2026-08-04T20:36:11Z","timestamp":1785875771423,"version":"3.56.0"},"reference-count":20,"publisher":"MDPI AG","issue":"4","license":[{"start":{"date-parts":[[2022,10,18]],"date-time":"2022-10-18T00:00:00Z","timestamp":1666051200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"3IA C\u00f4te d\u2019Azur project","award":["ANR-19-P3IA-0002"],"award-info":[{"award-number":["ANR-19-P3IA-0002"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["JCP"],"abstract":"<jats:p>This work studies the power of adversarial attacks against machine learning algorithms that use differentially private mechanisms as their weapon. In our setting, the adversary aims to modify the content of a statistical dataset via insertion of additional data without being detected by using the differential privacy to her\/his own benefit. The goal of this study is to evaluate how easy it is to detect such attacks (anomalies) when the adversary makes use of Gaussian and Laplacian perturbation using both statistical and information-theoretic tools. To this end, firstly via hypothesis testing, we characterize statistical thresholds for the adversary in various settings, which balances the privacy budget and the impact of the attack (the modification applied on the original data) in order to avoid being detected. In addition, we establish the privacy-distortion trade-off in the sense of the well-known rate-distortion function for the Gaussian mechanism by using an information-theoretic approach. Accordingly, we derive an upper bound on the variance of the attacker\u2019s additional data as a function of the sensitivity and the original data\u2019s second-order statistics. Lastly, we introduce a new privacy metric based on Chernoff information for anomaly detection under differential privacy as a stronger alternative for the (\u03f5,\u03b4)-differential privacy in Gaussian mechanisms. Analytical results are supported by numerical evaluations.<\/jats:p>","DOI":"10.3390\/jcp2040042","type":"journal-article","created":{"date-parts":[[2022,10,18]],"date-time":"2022-10-18T00:31:01Z","timestamp":1666053061000},"page":"830-852","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":4,"title":["Calibrating the Attack to Sensitivity in Differentially Private Mechanisms"],"prefix":"10.3390","volume":"2","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-5313-1168","authenticated-orcid":false,"given":"Ay\u015fe","family":"\u00dcnsal","sequence":"first","affiliation":[{"name":"Digital Security Department, EURECOM, Campus SophiaTECH, 450 Route des Chappes, 06140 Biot, France"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0269-9495","authenticated-orcid":false,"given":"Melek","family":"\u00d6nen","sequence":"additional","affiliation":[{"name":"Digital Security Department, EURECOM, Campus SophiaTECH, 450 Route des Chappes, 06140 Biot, France"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2022,10,18]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","unstructured":"Dwork, C. (2006, January 10\u201314). Differential Privacy. Proceedings of the Automata, Languages and Programming, Venice, Italy.","DOI":"10.1007\/11787006_1"},{"key":"ref_2","unstructured":"Dwork, C. (2008, January 25\u201329). Differential Privacy: A Survey of Results. Proceedings of the International Conference on Theory and Applications of Models of Computation TAMC 2008, LNCS 4978, Xi\u2019an, China."},{"key":"ref_3","first-page":"135","article-title":"Differential Privacy for Statistics: What we Know and What we Want to Learn","volume":"1","author":"Dwork","year":"2010","journal-title":"J. Priv. Confid."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"211","DOI":"10.1561\/0400000042","article-title":"The Algorithmic Foundations of Differential Privacy","volume":"9","author":"Dwork","year":"2014","journal-title":"Found. Trends Theor. Comput. Sci."},{"key":"ref_5","doi-asserted-by":"crossref","unstructured":"Cuff, P., and Yu, L. (2016, January 24\u201328). Differential Privacy as a Mutual Information Constraint. Proceedings of the CCS 2016\u201423rd ACM Conference on Computer and Communication Security, Vienna, Austria.","DOI":"10.1145\/2976749.2978308"},{"key":"ref_6","doi-asserted-by":"crossref","unstructured":"Giraldo, J., Cardenas, A.A., Kantarcioglu, M., and Katz, J. (2020, January 23\u201326). Adversarial Classification Under Differential Privacy. Proceedings of the NDSS 2020, Network and Distributed Systems Security Symposium, San Diego, CA, USA.","DOI":"10.14722\/ndss.2020.23047"},{"key":"ref_7","doi-asserted-by":"crossref","unstructured":"Dwork, C., McSherry, F., Nissim, K., and Smith, A. (2006, January 4\u20137). Calibrating Noise to Sensitivity in Private Data Analysis. Proceedings of the Theory of Cryptography Conference, New York, NY, USA.","DOI":"10.1007\/11681878_14"},{"key":"ref_8","doi-asserted-by":"crossref","unstructured":"Liu, C., He, X., Chanyaswad, T., Wang, S., and Mittal, P. (2019, January 16\u201320). Investigating Statistical Privacy Frameworks from the Perspective of Hypothesis Testing. Proceedings of the PETS 2019\u2014Privacy Enhancing Technologies, Stockholm, Sweden.","DOI":"10.2478\/popets-2019-0045"},{"key":"ref_9","unstructured":"Sheffet, O. (2018, January 14\u201316). Locally Private Hypothesis Testing. Proceedings of the Machine Learning Research, Beijing, China."},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"5018","DOI":"10.1109\/TIT.2016.2584610","article-title":"On the Relation Between Identifiability, Differential Privacy and Mutual Information Privacy","volume":"62","author":"Wang","year":"2016","journal-title":"IEEE Trans. Inf. Theory"},{"key":"ref_11","doi-asserted-by":"crossref","unstructured":"Sarwate, A., and Sankar, L. (2014, January 1\u20133). A Rate-Distortion Perspective on Local Differential Privacy. Proceedings of the Fiftieth Annual Allerton Conference, Monticello, IL, USA.","DOI":"10.1109\/ALLERTON.2014.7028550"},{"key":"ref_12","doi-asserted-by":"crossref","unstructured":"du Pin Calmon, F., and Fawaz, N. (2012, January 1\u20135). Privacy against Statistical Inference. Proceedings of the Fiftieth Annual Allerton Conference, Monticello, IL, USA.","DOI":"10.1109\/Allerton.2012.6483382"},{"key":"ref_13","first-page":"1","article-title":"Locally Differentially Private Randomized Response for Discrete Distribution Learning","volume":"22","author":"Pastore","year":"2021","journal-title":"J. Mach. Learn. Res."},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"133","DOI":"10.1016\/j.ijar.2019.07.003","article-title":"Adversarial Classification: An adversarial risk analysis","volume":"113","author":"Naveiro","year":"2019","journal-title":"Int. J. Approx. Reason."},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"841","DOI":"10.1198\/jasa.2009.0155","article-title":"Adversarial Risk Analysis","volume":"104","author":"Insua","year":"2009","journal-title":"J. Am. Stat. Assoc."},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Lowd, D., and Meek, C. (2005, January 21\u201324). Adversarial Learning. Proceedings of the Eleventh ACM SIGKDD International Conference on Knowledge Discovery in Data Mining KDD\u201905, Chicago, IL, USA.","DOI":"10.1145\/1081870.1081950"},{"key":"ref_17","doi-asserted-by":"crossref","unstructured":"Dalvi, N., Domingos, P., Sanghai, S., and Verma, D. (2004, January 22\u201325). Adversarial Classification. Proceedings of the Tenth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD\u201904, Seattle, WA, USA.","DOI":"10.1145\/1014052.1014066"},{"key":"ref_18","first-page":"289","article-title":"On the Problem of the Most Efficient Tests of Statistical Hypotheses","volume":"231","author":"Neyman","year":"1933","journal-title":"Philos. Trans. R. Soc. A"},{"key":"ref_19","unstructured":"Cover, T., and Thomas, J.A. (1991). Elements of Information Theory, Wiley."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"124","DOI":"10.1016\/j.ins.2013.06.018","article-title":"Renyi Divergence Measures for Commonly Used Univariate Continuous Distributions","volume":"249","author":"Gil","year":"2013","journal-title":"Inf. Sci."}],"container-title":["Journal of Cybersecurity and Privacy"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2624-800X\/2\/4\/42\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,11]],"date-time":"2025-10-11T00:56:07Z","timestamp":1760144167000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2624-800X\/2\/4\/42"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,10,18]]},"references-count":20,"journal-issue":{"issue":"4","published-online":{"date-parts":[[2022,12]]}},"alternative-id":["jcp2040042"],"URL":"https:\/\/doi.org\/10.3390\/jcp2040042","relation":{},"ISSN":["2624-800X"],"issn-type":[{"value":"2624-800X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,10,18]]}}}