FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2894553193> ?p ?o ?g. }

• W2894553193 abstract "We consider a model of robust learning in an adversarial environment. The learner gets uncorrupted training data with access to possible corruptions that may be effected by the adversary during testing. The learner's goal is to build a robust classifier, which will be tested on future adversarial examples. The adversary is limited to $k$ possible corruptions for each input. We model the learner-adversary interaction as a zero-sum game. This model is closely related to the adversarial examples model of Schmidt et al. (2018); Madry et al. (2017). Our main results consist of generalization bounds for the binary and multiclass classification, as well as the real-valued case (regression). For the binary classification setting, we both tighten the generalization bound of Feige, Mansour, and Schapire (2015), and are also able to handle infinite hypothesis classes. The sample complexity is improved from $O(frac{1}{epsilon^4}log(frac{|mathcal{H}|}{delta}))$ to $Obig(frac{1}{epsilon^2}(sqrt{k mathrm{VC}(mathcal{H})}log^{frac{3}{2}+alpha}(kmathrm{VC}(mathcal{H}))+log(frac{1}{delta})big)$ for any $alpha > 0$. Additionally, we extend the algorithm and generalization bound from the binary to the multiclass and real-valued cases. Along the way, we obtain results on fat-shattering dimension and Rademacher complexity of $k$-fold maxima over function classes; these may be of independent interest. For binary classification, the algorithm of Feige et al. (2015) uses a regret minimization algorithm and an ERM oracle as a black box; we adapt it for the multiclass and regression settings. The algorithm provides us with near-optimal policies for the players on a given training sample." @default.
• W2894553193 created "2018-10-12" @default.
• W2894553193 creator A5001138637 @default.
• W2894553193 creator A5014637159 @default.
• W2894553193 creator A5019371046 @default.
• W2894553193 date "2018-10-04" @default.
• W2894553193 modified "2023-09-27" @default.
• W2894553193 title "Improved generalization bounds for robust learning" @default.
• W2894553193 cites W1658064490 @default.
• W2894553193 cites W1673923490 @default.
• W2894553193 cites W1735424305 @default.
• W2894553193 cites W1766442844 @default.
• W2894553193 cites W1892947258 @default.
• W2894553193 cites W1905905554 @default.
• W2894553193 cites W1944672 @default.
• W2894553193 cites W1968998685 @default.
• W2894553193 cites W2005299943 @default.
• W2894553193 cites W2005951230 @default.
• W2894553193 cites W2014902932 @default.
• W2894553193 cites W2025263811 @default.
• W2894553193 cites W2071570468 @default.
• W2894553193 cites W2092074494 @default.
• W2894553193 cites W2093294140 @default.
• W2894553193 cites W2106887613 @default.
• W2894553193 cites W2107189314 @default.
• W2894553193 cites W2114296159 @default.
• W2894553193 cites W2142992973 @default.
• W2894553193 cites W2154952480 @default.
• W2894553193 cites W2243397390 @default.
• W2894553193 cites W2552767274 @default.
• W2894553193 cites W2579923771 @default.
• W2894553193 cites W2611675901 @default.
• W2894553193 cites W2612637113 @default.
• W2894553193 cites W2640329709 @default.
• W2894553193 cites W2767075075 @default.
• W2894553193 cites W2799031776 @default.
• W2894553193 cites W2803732607 @default.
• W2894553193 cites W2883441508 @default.
• W2894553193 cites W2892179671 @default.
• W2894553193 cites W2897297967 @default.
• W2894553193 cites W2906989593 @default.
• W2894553193 cites W2913450167 @default.
• W2894553193 cites W2947294642 @default.
• W2894553193 cites W2947421958 @default.
• W2894553193 cites W2950787360 @default.
• W2894553193 cites W2952049331 @default.
• W2894553193 cites W2962730916 @default.
• W2894553193 cites W2962902183 @default.
• W2894553193 cites W2962972504 @default.
• W2894553193 cites W2963207607 @default.
• W2894553193 cites W2963289726 @default.
• W2894553193 cites W2963353335 @default.
• W2894553193 cites W2963466683 @default.
• W2894553193 cites W2963777481 @default.
• W2894553193 cites W2964556751 @default.
• W2894553193 cites W2970203973 @default.
• W2894553193 cites W2970273067 @default.
• W2894553193 cites W2970433861 @default.
• W2894553193 cites W2970675290 @default.
• W2894553193 cites W2970680991 @default.
• W2894553193 cites W2970780203 @default.
• W2894553193 cites W2971316968 @default.
• W2894553193 cites W2989105640 @default.
• W2894553193 cites W2990565414 @default.
• W2894553193 cites W2994164853 @default.
• W2894553193 cites W9657784 @default.
• W2894553193 cites W1148694266 @default.
• W2894553193 hasPublicationYear "2018" @default.
• W2894553193 type Work @default.
• W2894553193 sameAs 2894553193 @default.
• W2894553193 citedByCount "10" @default.
• W2894553193 countsByYear W28945531932018 @default.
• W2894553193 countsByYear W28945531932019 @default.
• W2894553193 countsByYear W28945531932020 @default.
• W2894553193 countsByYear W28945531932021 @default.
• W2894553193 crossrefType "posted-content" @default.
• W2894553193 hasAuthorship W2894553193A5001138637 @default.
• W2894553193 hasAuthorship W2894553193A5014637159 @default.
• W2894553193 hasAuthorship W2894553193A5019371046 @default.
• W2894553193 hasConcept C11413529 @default.
• W2894553193 hasConcept C114614502 @default.
• W2894553193 hasConcept C118615104 @default.
• W2894553193 hasConcept C12267149 @default.
• W2894553193 hasConcept C134306372 @default.
• W2894553193 hasConcept C154945302 @default.
• W2894553193 hasConcept C177148314 @default.
• W2894553193 hasConcept C33923547 @default.
• W2894553193 hasConcept C37736160 @default.
• W2894553193 hasConcept C41008148 @default.
• W2894553193 hasConcept C48372109 @default.
• W2894553193 hasConcept C66905080 @default.
• W2894553193 hasConcept C77553402 @default.
• W2894553193 hasConcept C94375191 @default.
• W2894553193 hasConceptScore W2894553193C11413529 @default.
• W2894553193 hasConceptScore W2894553193C114614502 @default.
• W2894553193 hasConceptScore W2894553193C118615104 @default.
• W2894553193 hasConceptScore W2894553193C12267149 @default.
• W2894553193 hasConceptScore W2894553193C134306372 @default.
• W2894553193 hasConceptScore W2894553193C154945302 @default.
• W2894553193 hasConceptScore W2894553193C177148314 @default.

• next