FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4386746485> ?p ?o ?g. }

• W4386746485 endingPage "171" @default.
• W4386746485 startingPage "152" @default.
• W4386746485 abstract "Neural abstractions have been recently introduced as formal approximations of complex, nonlinear dynamical models. They comprise a neural ODE and a certified upper bound on the error between the abstract neural network and the concrete dynamical model. So far neural abstractions have exclusively been obtained as neural networks consisting entirely of $$textrm{ReLU}$$ activation functions, resulting in neural ODE models that have piecewise affine dynamics, and which can be equivalently interpreted as linear hybrid automata. In this work, we observe that the utility of an abstraction depends on its use: some scenarios might require coarse abstractions that are easier to analyse, whereas others might require more complex, refined abstractions. We therefore consider neural abstractions of alternative shapes, namely either piecewise constant or nonlinear non-polynomial (specifically, obtained via sigmoidal activations). We employ formal inductive synthesis procedures to generate neural abstractions that result in dynamical models with these semantics. Empirically, we demonstrate the trade-off that these different neural abstraction templates have vis-a-vis their precision and synthesis time, as well as the time required for their safety verification (done via reachability computation). We improve existing synthesis techniques to enable abstraction of higher-dimensional models, and additionally discuss the abstraction of complex neural ODEs to improve the efficiency of reachability analysis for these models." @default.
• W4386746485 created "2023-09-15" @default.
• W4386746485 creator A5036605932 @default.
• W4386746485 creator A5058973098 @default.
• W4386746485 creator A5091718585 @default.
• W4386746485 date "2023-01-01" @default.
• W4386746485 modified "2023-09-27" @default.
• W4386746485 title "On the Trade-Off Between Efficiency and Precision of Neural Abstraction" @default.
• W4386746485 cites W1480909796 @default.
• W4386746485 cites W1541882523 @default.
• W4386746485 cites W1547728445 @default.
• W4386746485 cites W1594130515 @default.
• W4386746485 cites W1598855039 @default.
• W4386746485 cites W172589331 @default.
• W4386746485 cites W1796008344 @default.
• W4386746485 cites W1979197687 @default.
• W4386746485 cites W2004463571 @default.
• W4386746485 cites W2015863039 @default.
• W4386746485 cites W2019750884 @default.
• W4386746485 cites W2024772987 @default.
• W4386746485 cites W2027793965 @default.
• W4386746485 cites W2060095702 @default.
• W4386746485 cites W2084859862 @default.
• W4386746485 cites W2099719845 @default.
• W4386746485 cites W2117218135 @default.
• W4386746485 cites W2124778357 @default.
• W4386746485 cites W2126221728 @default.
• W4386746485 cites W2147936671 @default.
• W4386746485 cites W2152533774 @default.
• W4386746485 cites W2154679417 @default.
• W4386746485 cites W2158395308 @default.
• W4386746485 cites W2158451642 @default.
• W4386746485 cites W2166280160 @default.
• W4386746485 cites W2176215692 @default.
• W4386746485 cites W2259853674 @default.
• W4386746485 cites W2292454014 @default.
• W4386746485 cites W2342015760 @default.
• W4386746485 cites W2492473950 @default.
• W4386746485 cites W2526694582 @default.
• W4386746485 cites W2568852336 @default.
• W4386746485 cites W2602950602 @default.
• W4386746485 cites W2605819566 @default.
• W4386746485 cites W2739936107 @default.
• W4386746485 cites W2757091317 @default.
• W4386746485 cites W2783515600 @default.
• W4386746485 cites W2938701611 @default.
• W4386746485 cites W2963334980 @default.
• W4386746485 cites W2980176594 @default.
• W4386746485 cites W2980176926 @default.
• W4386746485 cites W3008793775 @default.
• W4386746485 cites W3046272654 @default.
• W4386746485 cites W3080908617 @default.
• W4386746485 cites W3112385784 @default.
• W4386746485 cites W3115390673 @default.
• W4386746485 cites W3165816975 @default.
• W4386746485 cites W3184441520 @default.
• W4386746485 cites W3186194270 @default.
• W4386746485 cites W3199281236 @default.
• W4386746485 cites W4252959020 @default.
• W4386746485 cites W4283802686 @default.
• W4386746485 cites W4283820925 @default.
• W4386746485 cites W4293399525 @default.
• W4386746485 cites W856973001 @default.
• W4386746485 doi "https://doi.org/10.1007/978-3-031-43835-6_12" @default.
• W4386746485 hasPublicationYear "2023" @default.
• W4386746485 type Work @default.
• W4386746485 citedByCount "0" @default.
• W4386746485 crossrefType "book-chapter" @default.
• W4386746485 hasAuthorship W4386746485A5036605932 @default.
• W4386746485 hasAuthorship W4386746485A5058973098 @default.
• W4386746485 hasAuthorship W4386746485A5091718585 @default.
• W4386746485 hasConcept C111472728 @default.
• W4386746485 hasConcept C121332964 @default.
• W4386746485 hasConcept C124304363 @default.
• W4386746485 hasConcept C134306372 @default.
• W4386746485 hasConcept C136643341 @default.
• W4386746485 hasConcept C138885662 @default.
• W4386746485 hasConcept C154945302 @default.
• W4386746485 hasConcept C164660894 @default.
• W4386746485 hasConcept C28826006 @default.
• W4386746485 hasConcept C33923547 @default.
• W4386746485 hasConcept C34862557 @default.
• W4386746485 hasConcept C41008148 @default.
• W4386746485 hasConcept C50644808 @default.
• W4386746485 hasConcept C62520636 @default.
• W4386746485 hasConcept C79379906 @default.
• W4386746485 hasConcept C80444323 @default.
• W4386746485 hasConceptScore W4386746485C111472728 @default.
• W4386746485 hasConceptScore W4386746485C121332964 @default.
• W4386746485 hasConceptScore W4386746485C124304363 @default.
• W4386746485 hasConceptScore W4386746485C134306372 @default.
• W4386746485 hasConceptScore W4386746485C136643341 @default.
• W4386746485 hasConceptScore W4386746485C138885662 @default.
• W4386746485 hasConceptScore W4386746485C154945302 @default.
• W4386746485 hasConceptScore W4386746485C164660894 @default.
• W4386746485 hasConceptScore W4386746485C28826006 @default.
• W4386746485 hasConceptScore W4386746485C33923547 @default.
• W4386746485 hasConceptScore W4386746485C34862557 @default.

• next