FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4385660112> ?p ?o ?g. }

• W4385660112 endingPage "e1454" @default.
• W4385660112 startingPage "e1454" @default.
• W4385660112 abstract "When using deep learning models, one of the most critical vulnerabilities is their exposure to adversarial inputs, which can cause wrong decisions (e.g., incorrect classification of an image) with minor perturbations. To address this vulnerability, it becomes necessary to retrain the affected model against adversarial inputs as part of the software testing process. In order to make this process energy efficient, data scientists need support on which are the best guidance metrics for reducing the adversarial inputs to create and use during testing, as well as optimal dataset configurations.We examined six guidance metrics for retraining deep learning models, specifically with convolutional neural network architecture, and three retraining configurations. Our goal is to improve the convolutional neural networks against the attack of adversarial inputs with regard to the accuracy, resource utilization and execution time from the point of view of a data scientist in the context of image classification.We conducted an empirical study using five datasets for image classification. We explore: (a) the accuracy, resource utilization, and execution time of retraining convolutional neural networks with the guidance of six different guidance metrics (neuron coverage, likelihood-based surprise adequacy, distance-based surprise adequacy, DeepGini, softmax entropy and random), (b) the accuracy and resource utilization of retraining convolutional neural networks with three different configurations (one-step adversarial retraining, adversarial retraining and adversarial fine-tuning).We reveal that adversarial retraining from original model weights, and by ordering with uncertainty metrics, gives the best model w.r.t. accuracy, resource utilization, and execution time.Although more studies are necessary, we recommend data scientists use the above configuration and metrics to deal with the vulnerability to adversarial inputs of deep learning models, as they can improve their models against adversarial inputs without using many inputs and without creating numerous adversarial inputs. We also show that dataset size has an important impact on the results." @default.
• W4385660112 created "2023-08-09" @default.
• W4385660112 creator A5016544041 @default.
• W4385660112 creator A5027686521 @default.
• W4385660112 creator A5050955846 @default.
• W4385660112 creator A5055476405 @default.
• W4385660112 date "2023-08-08" @default.
• W4385660112 modified "2023-09-30" @default.
• W4385660112 title "Guiding the retraining of convolutional neural networks against adversarial inputs" @default.
• W4385660112 cites W1995875735 @default.
• W4385660112 cites W2041713059 @default.
• W4385660112 cites W2067713319 @default.
• W4385660112 cites W2112796928 @default.
• W4385660112 cites W2147800946 @default.
• W4385660112 cites W2954855426 @default.
• W4385660112 cites W2963327228 @default.
• W4385660112 cites W2963542245 @default.
• W4385660112 cites W2997532515 @default.
• W4385660112 cites W3030364939 @default.
• W4385660112 cites W3035584216 @default.
• W4385660112 cites W3036286896 @default.
• W4385660112 cites W3100321043 @default.
• W4385660112 cites W3111523098 @default.
• W4385660112 cites W3120991880 @default.
• W4385660112 cites W3121083082 @default.
• W4385660112 cites W3165066932 @default.
• W4385660112 cites W3197049923 @default.
• W4385660112 cites W3214897310 @default.
• W4385660112 cites W3216686652 @default.
• W4385660112 cites W4220891756 @default.
• W4385660112 cites W4285579917 @default.
• W4385660112 cites W4286957040 @default.
• W4385660112 cites W4287272378 @default.
• W4385660112 cites W4287274254 @default.
• W4385660112 cites W4287637349 @default.
• W4385660112 cites W4293580221 @default.
• W4385660112 cites W4293584023 @default.
• W4385660112 cites W4297814571 @default.
• W4385660112 cites W4306780108 @default.
• W4385660112 cites W4362508322 @default.
• W4385660112 doi "https://doi.org/10.7717/peerj-cs.1454" @default.
• W4385660112 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/37705636" @default.
• W4385660112 hasPublicationYear "2023" @default.
• W4385660112 type Work @default.
• W4385660112 citedByCount "0" @default.
• W4385660112 crossrefType "journal-article" @default.
• W4385660112 hasAuthorship W4385660112A5016544041 @default.
• W4385660112 hasAuthorship W4385660112A5027686521 @default.
• W4385660112 hasAuthorship W4385660112A5050955846 @default.
• W4385660112 hasAuthorship W4385660112A5055476405 @default.
• W4385660112 hasBestOaLocation W43856601121 @default.
• W4385660112 hasConcept C108583219 @default.
• W4385660112 hasConcept C111919701 @default.
• W4385660112 hasConcept C119857082 @default.
• W4385660112 hasConcept C124101348 @default.
• W4385660112 hasConcept C144133560 @default.
• W4385660112 hasConcept C153083717 @default.
• W4385660112 hasConcept C154945302 @default.
• W4385660112 hasConcept C155202549 @default.
• W4385660112 hasConcept C188441871 @default.
• W4385660112 hasConcept C2778712577 @default.
• W4385660112 hasConcept C37736160 @default.
• W4385660112 hasConcept C41008148 @default.
• W4385660112 hasConcept C50644808 @default.
• W4385660112 hasConcept C81363708 @default.
• W4385660112 hasConcept C98045186 @default.
• W4385660112 hasConceptScore W4385660112C108583219 @default.
• W4385660112 hasConceptScore W4385660112C111919701 @default.
• W4385660112 hasConceptScore W4385660112C119857082 @default.
• W4385660112 hasConceptScore W4385660112C124101348 @default.
• W4385660112 hasConceptScore W4385660112C144133560 @default.
• W4385660112 hasConceptScore W4385660112C153083717 @default.
• W4385660112 hasConceptScore W4385660112C154945302 @default.
• W4385660112 hasConceptScore W4385660112C155202549 @default.
• W4385660112 hasConceptScore W4385660112C188441871 @default.
• W4385660112 hasConceptScore W4385660112C2778712577 @default.
• W4385660112 hasConceptScore W4385660112C37736160 @default.
• W4385660112 hasConceptScore W4385660112C41008148 @default.
• W4385660112 hasConceptScore W4385660112C50644808 @default.
• W4385660112 hasConceptScore W4385660112C81363708 @default.
• W4385660112 hasConceptScore W4385660112C98045186 @default.
• W4385660112 hasLocation W43856601121 @default.
• W4385660112 hasLocation W43856601122 @default.
• W4385660112 hasLocation W43856601123 @default.
• W4385660112 hasOpenAccess W4385660112 @default.
• W4385660112 hasPrimaryLocation W43856601121 @default.
• W4385660112 hasRelatedWork W2731899572 @default.
• W4385660112 hasRelatedWork W2913997398 @default.
• W4385660112 hasRelatedWork W2977314777 @default.
• W4385660112 hasRelatedWork W2981169515 @default.
• W4385660112 hasRelatedWork W2999805992 @default.
• W4385660112 hasRelatedWork W3116150086 @default.
• W4385660112 hasRelatedWork W3133861977 @default.
• W4385660112 hasRelatedWork W4200173597 @default.
• W4385660112 hasRelatedWork W4312417841 @default.
• W4385660112 hasRelatedWork W4321369474 @default.
• W4385660112 hasVolume "9" @default.
• W4385660112 isParatext "false" @default.

• next