Matches in SemOpenAlex for { <https://semopenalex.org/work/W3112679672> ?p ?o ?g. }
Showing items 1 to 88 of
88
with 100 items per page.
- W3112679672 abstract "In many practical applications, deep neural networks have been typically deployed to operate as a black box predictor. Despite the high amount of work on interpretability and high demand on the reliability of these systems, they typically still have to include a human actor in the loop, to validate the decisions and handle unpredictable failures and unexpected corner cases. This is true in particular for failure-critical application domains, such as medical diagnosis. We present a novel approach to explain and support an interpretation of the decision-making process to a human expert operating a deep learning system based on Convolutional Neural Network (CNN). By modeling activation statistics on selected layers of a trained CNN via Gaussian Mixture Models (GMM), we develop a novel perceptual code in binary vector space that describes how the input sample is processed by the CNN. By measuring distances between pairs of samples in this perceptual encoding space, for any new input sample, we can now retrieve a set of most perceptually similar and dissimilar samples from an existing atlas of labeled samples, to support and clarify the decision made by the CNN model. Possible uses of this approach include for example Computer-Aided Diagnosis (CAD) systems working with medical imaging data, such as Magnetic Resonance Imaging (MRI) or Computed Tomography (CT) scans. We demonstrate the viability of our method in the domain of medical imaging for patient condition diagnosis, as the proposed decision explanation method via similar ground truth domain examples (e.g. from existing diagnosis archives) will be interpretable by the operating medical personnel. Our results indicate that our method is capable of detecting distinct prediction strategies that enable us to identify the most similar predictions from an existing atlas." @default.
- W3112679672 created "2020-12-21" @default.
- W3112679672 creator A5030064907 @default.
- W3112679672 creator A5042086801 @default.
- W3112679672 creator A5054838130 @default.
- W3112679672 date "2020-12-03" @default.
- W3112679672 modified "2023-09-27" @default.
- W3112679672 title "Explaining Predictions of Deep Neural Classifier via Activation Analysis." @default.
- W3112679672 cites W1787224781 @default.
- W3112679672 cites W1996796871 @default.
- W3112679672 cites W2049633694 @default.
- W3112679672 cites W2133287637 @default.
- W3112679672 cites W2753415590 @default.
- W3112679672 cites W2765813195 @default.
- W3112679672 cites W2778796877 @default.
- W3112679672 cites W2890457888 @default.
- W3112679672 cites W2892375480 @default.
- W3112679672 cites W2905307056 @default.
- W3112679672 cites W2915292867 @default.
- W3112679672 cites W2949780682 @default.
- W3112679672 cites W2965914043 @default.
- W3112679672 cites W2973007151 @default.
- W3112679672 cites W3003663127 @default.
- W3112679672 hasPublicationYear "2020" @default.
- W3112679672 type Work @default.
- W3112679672 sameAs 3112679672 @default.
- W3112679672 citedByCount "0" @default.
- W3112679672 crossrefType "posted-content" @default.
- W3112679672 hasAuthorship W3112679672A5030064907 @default.
- W3112679672 hasAuthorship W3112679672A5042086801 @default.
- W3112679672 hasAuthorship W3112679672A5054838130 @default.
- W3112679672 hasConcept C108583219 @default.
- W3112679672 hasConcept C11413529 @default.
- W3112679672 hasConcept C119857082 @default.
- W3112679672 hasConcept C142724271 @default.
- W3112679672 hasConcept C153180895 @default.
- W3112679672 hasConcept C154945302 @default.
- W3112679672 hasConcept C2781067378 @default.
- W3112679672 hasConcept C31601959 @default.
- W3112679672 hasConcept C3309909 @default.
- W3112679672 hasConcept C41008148 @default.
- W3112679672 hasConcept C50644808 @default.
- W3112679672 hasConcept C534262118 @default.
- W3112679672 hasConcept C71924100 @default.
- W3112679672 hasConcept C81363708 @default.
- W3112679672 hasConcept C95623464 @default.
- W3112679672 hasConceptScore W3112679672C108583219 @default.
- W3112679672 hasConceptScore W3112679672C11413529 @default.
- W3112679672 hasConceptScore W3112679672C119857082 @default.
- W3112679672 hasConceptScore W3112679672C142724271 @default.
- W3112679672 hasConceptScore W3112679672C153180895 @default.
- W3112679672 hasConceptScore W3112679672C154945302 @default.
- W3112679672 hasConceptScore W3112679672C2781067378 @default.
- W3112679672 hasConceptScore W3112679672C31601959 @default.
- W3112679672 hasConceptScore W3112679672C3309909 @default.
- W3112679672 hasConceptScore W3112679672C41008148 @default.
- W3112679672 hasConceptScore W3112679672C50644808 @default.
- W3112679672 hasConceptScore W3112679672C534262118 @default.
- W3112679672 hasConceptScore W3112679672C71924100 @default.
- W3112679672 hasConceptScore W3112679672C81363708 @default.
- W3112679672 hasConceptScore W3112679672C95623464 @default.
- W3112679672 hasLocation W31126796721 @default.
- W3112679672 hasOpenAccess W3112679672 @default.
- W3112679672 hasPrimaryLocation W31126796721 @default.
- W3112679672 hasRelatedWork W2585691031 @default.
- W3112679672 hasRelatedWork W270432131 @default.
- W3112679672 hasRelatedWork W2793701907 @default.
- W3112679672 hasRelatedWork W2950860813 @default.
- W3112679672 hasRelatedWork W2963081790 @default.
- W3112679672 hasRelatedWork W2972156365 @default.
- W3112679672 hasRelatedWork W2990201905 @default.
- W3112679672 hasRelatedWork W3016830346 @default.
- W3112679672 hasRelatedWork W3017445145 @default.
- W3112679672 hasRelatedWork W3030475683 @default.
- W3112679672 hasRelatedWork W3034860610 @default.
- W3112679672 hasRelatedWork W3035856201 @default.
- W3112679672 hasRelatedWork W3036146103 @default.
- W3112679672 hasRelatedWork W3053495551 @default.
- W3112679672 hasRelatedWork W3088222393 @default.
- W3112679672 hasRelatedWork W3095253297 @default.
- W3112679672 hasRelatedWork W3097501873 @default.
- W3112679672 hasRelatedWork W3163418296 @default.
- W3112679672 hasRelatedWork W3180112443 @default.
- W3112679672 hasRelatedWork W3198112749 @default.
- W3112679672 isParatext "false" @default.
- W3112679672 isRetracted "false" @default.
- W3112679672 magId "3112679672" @default.
- W3112679672 workType "article" @default.