Matches in SemOpenAlex for { <https://semopenalex.org/work/W3016892025> ?p ?o ?g. }
Showing items 1 to 97 of
97
with 100 items per page.
- W3016892025 endingPage "982" @default.
- W3016892025 startingPage "969" @default.
- W3016892025 abstract "Machine learning models have been widely used in security applications. However, it is well-known that adversaries can adapt their attacks to evade detection. There has been some work on making machine learning models more robust to such attacks. However, one simple but promising approach called <italic xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>randomization</i> is under-explored. In addition, most existing works focus on models with differentiable error functions while tree-based models do not have such error functions but are quite popular because they are easy to interpret. This paper proposes a novel randomization-based approach to improve robustness of tree-based models against evasion attacks. The proposed approach incorporates randomization into both model training time and model application time (meaning when the model is used to detect attacks). We also apply this approach to random forest, an existing ML method which already has incorporated randomness at training time but still often fails to generate robust models. We proposed a novel weighted-random-forest method to generate more robust models and a clustering method to add randomness at model application time. We also proposed a theoretical framework to provide a lower bound for adversaries’ effort. Experiments on intrusion detection and spam filtering data show that our approach further improves robustness of random-forest method." @default.
- W3016892025 created "2020-04-24" @default.
- W3016892025 creator A5006806882 @default.
- W3016892025 creator A5051208435 @default.
- W3016892025 creator A5074754590 @default.
- W3016892025 date "2022-02-01" @default.
- W3016892025 modified "2023-09-23" @default.
- W3016892025 title "Using Randomness to Improve Robustness of Tree-Based Models Against Evasion Attacks" @default.
- W3016892025 cites W1484010841 @default.
- W3016892025 cites W1965052658 @default.
- W3016892025 cites W1966912382 @default.
- W3016892025 cites W1968998685 @default.
- W3016892025 cites W2008224380 @default.
- W3016892025 cites W2015585187 @default.
- W3016892025 cites W2047237187 @default.
- W3016892025 cites W2077742958 @default.
- W3016892025 cites W2109300365 @default.
- W3016892025 cites W2109857912 @default.
- W3016892025 cites W2114296159 @default.
- W3016892025 cites W2122646361 @default.
- W3016892025 cites W2141029621 @default.
- W3016892025 cites W2144083192 @default.
- W3016892025 cites W2145901875 @default.
- W3016892025 cites W2146211964 @default.
- W3016892025 cites W2150973844 @default.
- W3016892025 cites W2151298633 @default.
- W3016892025 cites W2161470437 @default.
- W3016892025 cites W2162275200 @default.
- W3016892025 cites W2180612164 @default.
- W3016892025 cites W2293768274 @default.
- W3016892025 cites W2296452361 @default.
- W3016892025 cites W2535873859 @default.
- W3016892025 cites W2611675901 @default.
- W3016892025 cites W2962700793 @default.
- W3016892025 cites W2964082701 @default.
- W3016892025 cites W3123545922 @default.
- W3016892025 cites W4297970707 @default.
- W3016892025 cites W9657784 @default.
- W3016892025 doi "https://doi.org/10.1109/tkde.2020.2987299" @default.
- W3016892025 hasPublicationYear "2022" @default.
- W3016892025 type Work @default.
- W3016892025 sameAs 3016892025 @default.
- W3016892025 citedByCount "5" @default.
- W3016892025 countsByYear W30168920252020 @default.
- W3016892025 countsByYear W30168920252021 @default.
- W3016892025 countsByYear W30168920252022 @default.
- W3016892025 countsByYear W30168920252023 @default.
- W3016892025 crossrefType "journal-article" @default.
- W3016892025 hasAuthorship W3016892025A5006806882 @default.
- W3016892025 hasAuthorship W3016892025A5051208435 @default.
- W3016892025 hasAuthorship W3016892025A5074754590 @default.
- W3016892025 hasConcept C104317684 @default.
- W3016892025 hasConcept C105795698 @default.
- W3016892025 hasConcept C119857082 @default.
- W3016892025 hasConcept C124101348 @default.
- W3016892025 hasConcept C125112378 @default.
- W3016892025 hasConcept C154945302 @default.
- W3016892025 hasConcept C169258074 @default.
- W3016892025 hasConcept C185592680 @default.
- W3016892025 hasConcept C33923547 @default.
- W3016892025 hasConcept C41008148 @default.
- W3016892025 hasConcept C55493867 @default.
- W3016892025 hasConcept C63479239 @default.
- W3016892025 hasConceptScore W3016892025C104317684 @default.
- W3016892025 hasConceptScore W3016892025C105795698 @default.
- W3016892025 hasConceptScore W3016892025C119857082 @default.
- W3016892025 hasConceptScore W3016892025C124101348 @default.
- W3016892025 hasConceptScore W3016892025C125112378 @default.
- W3016892025 hasConceptScore W3016892025C154945302 @default.
- W3016892025 hasConceptScore W3016892025C169258074 @default.
- W3016892025 hasConceptScore W3016892025C185592680 @default.
- W3016892025 hasConceptScore W3016892025C33923547 @default.
- W3016892025 hasConceptScore W3016892025C41008148 @default.
- W3016892025 hasConceptScore W3016892025C55493867 @default.
- W3016892025 hasConceptScore W3016892025C63479239 @default.
- W3016892025 hasFunder F4320337345 @default.
- W3016892025 hasIssue "2" @default.
- W3016892025 hasLocation W30168920251 @default.
- W3016892025 hasOpenAccess W3016892025 @default.
- W3016892025 hasPrimaryLocation W30168920251 @default.
- W3016892025 hasRelatedWork W2911455822 @default.
- W3016892025 hasRelatedWork W3018959556 @default.
- W3016892025 hasRelatedWork W3174196512 @default.
- W3016892025 hasRelatedWork W3211546796 @default.
- W3016892025 hasRelatedWork W4281560664 @default.
- W3016892025 hasRelatedWork W4281616679 @default.
- W3016892025 hasRelatedWork W4293525103 @default.
- W3016892025 hasRelatedWork W4308191010 @default.
- W3016892025 hasRelatedWork W4318350883 @default.
- W3016892025 hasRelatedWork W4323021782 @default.
- W3016892025 hasVolume "34" @default.
- W3016892025 isParatext "false" @default.
- W3016892025 isRetracted "false" @default.
- W3016892025 magId "3016892025" @default.
- W3016892025 workType "article" @default.