On Pixel-Wise Explanations for Non-Linear Classifier Decisions by Layer-Wise Relevance Propagation
Pourquoi ce travail est-il dans la base ?
Une base qui oublie comment elle a trouvé un travail ne peut pas être vérifiée. Voici les voies qui ont admis celui-ci.
Aucune affiliation canadienne. Une base fondée sur la seule affiliation (le devis habituel) n'aurait jamais vu ce travail. C'est l'un des travaux qui justifient l'inversion de la base.
Résumé
Understanding and interpreting classification decisions of automated image classification systems is of high value in many applications, as it allows to verify the reasoning of the system and provides additional information to the human expert. Although machine learning methods are solving very successfully a plethora of tasks, they have in most cases the disadvantage of acting as a black box, not providing any information about what made them arrive at a particular decision. This work proposes a general solution to the problem of understanding classification decisions by pixel-wise decomposition of nonlinear classifiers. We introduce a methodology that allows to visualize the contributions of single pixels to predictions for kernel-based classifiers over Bag of Words features and for multilayered neural networks. These pixel contributions can be visualized as heatmaps and are provided to a human expert who can intuitively not only verify the validity of the classification decision, but also focus further analysis on regions of potential interest. We evaluate our method for classifiers trained on PASCAL VOC 2009 images, synthetic image data containing geometric shapes, the MNIST handwritten digits data set and for the pre-trained ImageNet model available as part of the Caffe open source package.
Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.
La notice
- Revue
- PLoS ONE
- Thématique
- Advanced Neural Network Applications
- Domaine
- Computer Science
- Établissements canadiens
- —
- Organismes subventionnaires
- Technische Universität BerlinBundesministerium für Wirtschaft und TechnologieBundesministerium für Bildung und ForschungBanting and Best Diabetes Centre, University of TorontoNational Research Foundation of KoreaDeutsche ForschungsgemeinschaftNational Research Foundation
- Mots-clés
- MNIST databaseComputer scienceArtificial intelligencePixelPascal (unit)Machine learningPattern recognition (psychology)Classifier (UML)Contextual image classificationArtificial neural networkData miningImage (mathematics)
- Résumé présent dans OpenAlex
- oui