Matches in SemOpenAlex for { <https://semopenalex.org/work/W2994857703> ?p ?o ?g. }
Showing items 1 to 95 of
95
with 100 items per page.
- W2994857703 abstract "Neural Networks are being integrated into safety critical systems, e.g., perception systems for autonomous vehicles, which require trained networks to perform safely in novel scenarios. It is challenging to verify neural networks because their decisions are not explainable, they cannot be exhaustively tested, and finite test samples cannot capture the variation across all operating conditions. Existing work seeks to train models robust to new scenarios via domain adaptation, style transfer, or few-shot learning. But these techniques fail to predict how a trained model will perform when the operating conditions differ from the testing conditions. We propose a metric, Machine Learning (ML) Dependability, that measures the network's probability of success in specified operating conditions which need not be the testing conditions. In addition, we propose the metrics Task Undependability and Harmful Undependability to distinguish network failures by their consequences. We evaluate the performance of a Neural Network agent trained using Reinforcement Learning in a simulated robot manipulation task. Our results demonstrate that we can accurately predict the ML Dependability, Task Undependability, and Harmful Undependability for operating conditions that are significantly different from the testing conditions. Finally, we design a Safety Function, using harmful failures identified during testing, that reduces harmful failures, in one example, by a factor of 700 while maintaining a high probability of success." @default.
- W2994857703 created "2019-12-26" @default.
- W2994857703 creator A5025072260 @default.
- W2994857703 creator A5034146992 @default.
- W2994857703 creator A5046354096 @default.
- W2994857703 creator A5088325127 @default.
- W2994857703 date "2019-12-20" @default.
- W2994857703 modified "2023-09-28" @default.
- W2994857703 title "Dependable Neural Networks for Safety Critical Tasks" @default.
- W2994857703 cites W1556520454 @default.
- W2994857703 cites W1979345446 @default.
- W2994857703 cites W2513988836 @default.
- W2994857703 cites W2766734825 @default.
- W2994857703 cites W2799352588 @default.
- W2994857703 cites W2910453440 @default.
- W2994857703 cites W2913211631 @default.
- W2994857703 cites W2951262671 @default.
- W2994857703 cites W2962793481 @default.
- W2994857703 cites W2963789515 @default.
- W2994857703 cites W2963958573 @default.
- W2994857703 cites W2964277612 @default.
- W2994857703 cites W2972049650 @default.
- W2994857703 cites W2981207549 @default.
- W2994857703 hasPublicationYear "2019" @default.
- W2994857703 type Work @default.
- W2994857703 sameAs 2994857703 @default.
- W2994857703 citedByCount "0" @default.
- W2994857703 crossrefType "posted-content" @default.
- W2994857703 hasAuthorship W2994857703A5025072260 @default.
- W2994857703 hasAuthorship W2994857703A5034146992 @default.
- W2994857703 hasAuthorship W2994857703A5046354096 @default.
- W2994857703 hasAuthorship W2994857703A5088325127 @default.
- W2994857703 hasConcept C115903868 @default.
- W2994857703 hasConcept C119857082 @default.
- W2994857703 hasConcept C120665830 @default.
- W2994857703 hasConcept C121332964 @default.
- W2994857703 hasConcept C127413603 @default.
- W2994857703 hasConcept C134306372 @default.
- W2994857703 hasConcept C139807058 @default.
- W2994857703 hasConcept C154945302 @default.
- W2994857703 hasConcept C176217482 @default.
- W2994857703 hasConcept C200601418 @default.
- W2994857703 hasConcept C201995342 @default.
- W2994857703 hasConcept C21547014 @default.
- W2994857703 hasConcept C2780451532 @default.
- W2994857703 hasConcept C33923547 @default.
- W2994857703 hasConcept C36503486 @default.
- W2994857703 hasConcept C41008148 @default.
- W2994857703 hasConcept C50644808 @default.
- W2994857703 hasConcept C77019957 @default.
- W2994857703 hasConcept C97541855 @default.
- W2994857703 hasConceptScore W2994857703C115903868 @default.
- W2994857703 hasConceptScore W2994857703C119857082 @default.
- W2994857703 hasConceptScore W2994857703C120665830 @default.
- W2994857703 hasConceptScore W2994857703C121332964 @default.
- W2994857703 hasConceptScore W2994857703C127413603 @default.
- W2994857703 hasConceptScore W2994857703C134306372 @default.
- W2994857703 hasConceptScore W2994857703C139807058 @default.
- W2994857703 hasConceptScore W2994857703C154945302 @default.
- W2994857703 hasConceptScore W2994857703C176217482 @default.
- W2994857703 hasConceptScore W2994857703C200601418 @default.
- W2994857703 hasConceptScore W2994857703C201995342 @default.
- W2994857703 hasConceptScore W2994857703C21547014 @default.
- W2994857703 hasConceptScore W2994857703C2780451532 @default.
- W2994857703 hasConceptScore W2994857703C33923547 @default.
- W2994857703 hasConceptScore W2994857703C36503486 @default.
- W2994857703 hasConceptScore W2994857703C41008148 @default.
- W2994857703 hasConceptScore W2994857703C50644808 @default.
- W2994857703 hasConceptScore W2994857703C77019957 @default.
- W2994857703 hasConceptScore W2994857703C97541855 @default.
- W2994857703 hasOpenAccess W2994857703 @default.
- W2994857703 hasRelatedWork W1811890870 @default.
- W2994857703 hasRelatedWork W2058764642 @default.
- W2994857703 hasRelatedWork W2304007452 @default.
- W2994857703 hasRelatedWork W2346041701 @default.
- W2994857703 hasRelatedWork W2466170565 @default.
- W2994857703 hasRelatedWork W2472431160 @default.
- W2994857703 hasRelatedWork W2474284978 @default.
- W2994857703 hasRelatedWork W2522695573 @default.
- W2994857703 hasRelatedWork W2612112140 @default.
- W2994857703 hasRelatedWork W2754002858 @default.
- W2994857703 hasRelatedWork W2900218414 @default.
- W2994857703 hasRelatedWork W2903075639 @default.
- W2994857703 hasRelatedWork W2911724991 @default.
- W2994857703 hasRelatedWork W2966688529 @default.
- W2994857703 hasRelatedWork W2971143001 @default.
- W2994857703 hasRelatedWork W2975984998 @default.
- W2994857703 hasRelatedWork W2978309085 @default.
- W2994857703 hasRelatedWork W2989008014 @default.
- W2994857703 hasRelatedWork W2989745381 @default.
- W2994857703 hasRelatedWork W2991158008 @default.
- W2994857703 isParatext "false" @default.
- W2994857703 isRetracted "false" @default.
- W2994857703 magId "2994857703" @default.
- W2994857703 workType "article" @default.