Matches in SemOpenAlex for { <https://semopenalex.org/work/W3042715818> ?p ?o ?g. }
- W3042715818 abstract "The black-box nature of deep neural networks (DNNs) makes it impossible to understand why a particular output is produced, creating demand for Explainable AI. In this paper, we show that statistical fault localization (SFL) techniques from software engineering deliver high quality explanations of the outputs of DNNs, where we define an explanation as a minimal subset of features sufficient for making the same decision as for the original input. We present an algorithm and a tool called DeepCover, which synthesizes a ranking of the features of the inputs using SFL and constructs explanations for the decisions of the DNN based on this ranking. We compare explanations produced by DeepCover with those of the state-of-the-art tools GradCAM, LIME, SHAP, RISE and Extremal and show that explanations generated by DeepCover are consistently better across a broad set of experiments. On a benchmark set with known ground truth, DeepCover achieves 76.7% accuracy, which is 6% better than the second best Extremal." @default.
- W3042715818 created "2020-07-23" @default.
- W3042715818 creator A5020085889 @default.
- W3042715818 creator A5052794441 @default.
- W3042715818 creator A5053852359 @default.
- W3042715818 creator A5086206346 @default.
- W3042715818 date "2019-08-06" @default.
- W3042715818 modified "2023-09-27" @default.
- W3042715818 title "Explaining Image Classifiers using Statistical Fault Localization" @default.
- W3042715818 cites W1569027050 @default.
- W3042715818 cites W1686810756 @default.
- W3042715818 cites W1773421521 @default.
- W3042715818 cites W1787224781 @default.
- W3042715818 cites W1973644552 @default.
- W3042715818 cites W1981711716 @default.
- W3042715818 cites W2010833880 @default.
- W3042715818 cites W2024507356 @default.
- W3042715818 cites W2040577374 @default.
- W3042715818 cites W2052879098 @default.
- W3042715818 cites W2067416361 @default.
- W3042715818 cites W2094837707 @default.
- W3042715818 cites W2101819268 @default.
- W3042715818 cites W2116737258 @default.
- W3042715818 cites W2119136132 @default.
- W3042715818 cites W2131346202 @default.
- W3042715818 cites W2133728573 @default.
- W3042715818 cites W2137819283 @default.
- W3042715818 cites W2147699889 @default.
- W3042715818 cites W2167381349 @default.
- W3042715818 cites W2183341477 @default.
- W3042715818 cites W2242274110 @default.
- W3042715818 cites W2282821441 @default.
- W3042715818 cites W2308882098 @default.
- W3042715818 cites W2324042828 @default.
- W3042715818 cites W2335386552 @default.
- W3042715818 cites W2343875716 @default.
- W3042715818 cites W2467903332 @default.
- W3042715818 cites W2510508396 @default.
- W3042715818 cites W2531409750 @default.
- W3042715818 cites W2578469907 @default.
- W3042715818 cites W2593050976 @default.
- W3042715818 cites W2603808365 @default.
- W3042715818 cites W2605409611 @default.
- W3042715818 cites W2609398547 @default.
- W3042715818 cites W2612445135 @default.
- W3042715818 cites W2616028256 @default.
- W3042715818 cites W2753783305 @default.
- W3042715818 cites W2785685510 @default.
- W3042715818 cites W2792641098 @default.
- W3042715818 cites W2793633339 @default.
- W3042715818 cites W2802945484 @default.
- W3042715818 cites W2809136100 @default.
- W3042715818 cites W2809182766 @default.
- W3042715818 cites W2809701591 @default.
- W3042715818 cites W2850992922 @default.
- W3042715818 cites W2884199749 @default.
- W3042715818 cites W2884426148 @default.
- W3042715818 cites W2888510757 @default.
- W3042715818 cites W2888527474 @default.
- W3042715818 cites W2891612330 @default.
- W3042715818 cites W2901765484 @default.
- W3042715818 cites W2914605630 @default.
- W3042715818 cites W2933011641 @default.
- W3042715818 cites W2949197630 @default.
- W3042715818 cites W2949205117 @default.
- W3042715818 cites W2952527295 @default.
- W3042715818 cites W2962715466 @default.
- W3042715818 cites W2962858109 @default.
- W3042715818 cites W2962862931 @default.
- W3042715818 cites W2963207607 @default.
- W3042715818 cites W2963327228 @default.
- W3042715818 cites W2963412070 @default.
- W3042715818 cites W2963733622 @default.
- W3042715818 cites W2963857521 @default.
- W3042715818 cites W2963913218 @default.
- W3042715818 cites W2964153729 @default.
- W3042715818 cites W2988157455 @default.
- W3042715818 cites W3036286896 @default.
- W3042715818 cites W3089623150 @default.
- W3042715818 cites W3091157996 @default.
- W3042715818 cites W3103557498 @default.
- W3042715818 cites W3105599650 @default.
- W3042715818 cites W3118608800 @default.
- W3042715818 cites W2796430303 @default.
- W3042715818 doi "https://doi.org/10.48550/arxiv.1908.02374" @default.
- W3042715818 hasPublicationYear "2019" @default.
- W3042715818 type Work @default.
- W3042715818 sameAs 3042715818 @default.
- W3042715818 citedByCount "1" @default.
- W3042715818 countsByYear W30427158182021 @default.
- W3042715818 crossrefType "posted-content" @default.
- W3042715818 hasAuthorship W3042715818A5020085889 @default.
- W3042715818 hasAuthorship W3042715818A5052794441 @default.
- W3042715818 hasAuthorship W3042715818A5053852359 @default.
- W3042715818 hasAuthorship W3042715818A5086206346 @default.
- W3042715818 hasBestOaLocation W30427158181 @default.
- W3042715818 hasConcept C111472728 @default.
- W3042715818 hasConcept C115961682 @default.
- W3042715818 hasConcept C119857082 @default.
- W3042715818 hasConcept C124101348 @default.