05189nam 2200649Ii 45000010011000000030004000110050017000150060024000320070015000560080041000710200030001120200026001420240028001680400030001960500026002260820017002521000027002692450093002962640056003893000054004453360021004993370026005203380032005464900108005785040057006865050126007435050191008695050280010605050327013405050297016675050413019645050050023775050031024275060065024585100019025235100023025425100011025655100011025765100024025875100039026115100031026505100019026815100026027005201146027265240194038725300029040665380036040955380047041315880045041786500022042236500022042456550022042677100032042897760034043218300109043558560075044640100000102NOW20210210190106.0m eo d cr cn |||m|||a190401s2021 maua ob 000 0 eng d a9781680837650qelectronic z9781680837643q(print)7 a10.1561/01000001022doi aCaBNVSLcCaBNVSLdCaBNVSL 4aTK5102.5b.B37 2021eb04a621.38222231 aBarni, Mauro,eauthor.10aTheoretical foundations of adversarial binary detection /cMauro Barni, Benedetta Tondi. 1a[Hanover, Massachusetts] :bNow Publishers,c2021. a1 PDF (pages 1-172) :billustrations (some color) atext2rdacontent aelectronic2isbdmedia aonline resource2rdacarrier1 aFoundations and trends in communications and information theory,x1567-2328 ;vVol. 18: No. 1, pp 1-172 aIncludes bibliographical references (pages 164-172).0 a1. Introduction. 1.1. Application areas ; 1.2. Scope of the theory ; 1.3. Related work ; 1.4. Outline of the monograph --8 a2. Background notions and tools. 2.1. Notation and definitions ; 2.2. Game theory in a nutshell ; 2.3. Introduction to optimal transport (OT) ; 2.4. Elements of large deviation theory --8 a3. Binary detection game with known sources. 3.1. Detection game with known sources (DG-KS) ; 3.2. Solution of the DG-KS game ; 3.3. Analysis of the payoff at the equilibrium ; 3.4. Numerical analysis: a case study ; 3.5. DG-KS Game under maximum distortion-limited attack --0 a4. Limit performance and source distinguishability. 4.1. Characterization of the indistinguishability region using OT ; 4.2 Best achievable performance in the DG-KS setup ; 4.3. Security margin in the DG-KS setup ; 4.4. Security margin computation ; 4.5. Source distinguishability under maximumdistortion-limited attack --8 a5. Binary detection game with training data. 5.1. Detection game with training data (DG-TR) ; 5.2. Asymptotic equilibrium of the DG-TRb game ; 5.3. Analysis of the payoff at the equilibrium ; 5.4. Game with independent training sequences (DG-TRa) ; 5.5. Security margin in the DG-TR setup -- 8 a6. Binary detection games with corrupted training. 6.1. Discussion and link with adversarial machine learning ; 6.2 Detection game with corrupted training (DG-CTR) ; 6.3. The DG-CTRa game ; 6.4. Solution of the DG-CTRat and DG-CTRa games ; 6.5. Source distinguishability in the DG-CTRa setup ; 6.6. The DG-CTRr game ; 6.7. Solution of the DG-CTRr game ; 6.8. Source distinguishability in the DG-CTRr setup --8 a7. Summary and outlook -- Acknowledgements --8 aAppendices -- References. aRestricted to subscribers or individual document purchasers.0 aGoogle Scholar0 aGoogle Book Search0 aINSPEC0 aScopus0 aACM Computing Guide0 aDBLP Computer Science Bibliography0 aZentralblatt MATH Database0 aAMS MathSciNet0 aACM Computing Reviews3 aThe present monograph focuses on the detection problem in adversarial setting. When framed in an adversarial setting, classical detection theory can not be applied any more, since, in order to make a correct decision, the presence of an adversary must be taken into account when designing the detector. In particular, the interplay between the Defender (), wishing to carry out the detection task, and the Attacker (), aiming at impeding it, must be investigated. The purpose of this monograph is to lay out the foundations of a general theory of adversarial detection, taking into account the impact that the presence of the adversary has on the design of the optimal detector. We do so by casting the adversarial detection problem into a game theoretical framework, which is then studied by relying on typical methods of information theory. As a final result, the theory allows to state the conditions under which both the false positive and false negative error probabilities tend to zero exponentially fast, and to relate the error exponents of the two kinds of errors to the distortion the attacker can introduce into the test sequence. aMauro Barni and Benedetta Tondi (2020), "Theoretical Foundations of Adversarial Binary Detection", Foundations and Trends in Communications and Information Theory: Vol. 18: No. 1, pp 1-172. aAlso available in print. aMode of access: World Wide Web. aSystem requirements: Adobe Acrobat reader. aTitle from PDF (viewed on Feb. 8, 2021). 0aSignal detection. 0aSecurity systems. 0aElectronic books.2 aNow Publishers,epublisher.08iPrint version:z9781680837643 0aFoundations and trends in communications and information theory ;vVol. 18: No. 1, pp 1-172.x1567-2328 483Abstract with links to full textuhttp://dx.doi.org/10.1561/0100000102