Matches in SemOpenAlex for { <https://semopenalex.org/work/W3204118220> ?p ?o ?g. }
Showing items 1 to 82 of
82
with 100 items per page.
- W3204118220 abstract "We present a computational method for empirically characterizing the training loss level-sets of deep neural networks. Our method numerically constructs a path in parameter space that is constrained to a set with a fixed near-zero training loss. By measuring regularization functions and test loss at different points within this path, we examine how different points in the parameter space with the same fixed training loss compare in terms of generalization ability. We also compare this method for finding regularized points with the more typical method, that uses objective functions which are weighted sums of training loss and regularization terms. We apply dimensionality reduction to the traversed paths in order to visualize the loss level sets in a well-regularized region of parameter space. Our results provide new information about the loss landscape of deep neural networks, as well as a new strategy for reducing test loss." @default.
- W3204118220 created "2021-10-11" @default.
- W3204118220 creator A5051693199 @default.
- W3204118220 creator A5085686472 @default.
- W3204118220 date "2021-07-18" @default.
- W3204118220 modified "2023-10-14" @default.
- W3204118220 title "Numerical Exploration of Training Loss Level-Sets in Deep Neural Networks" @default.
- W3204118220 cites W1481566577 @default.
- W3204118220 cites W1533861849 @default.
- W3204118220 cites W2001619934 @default.
- W3204118220 cites W2086513155 @default.
- W3204118220 cites W2111072639 @default.
- W3204118220 cites W2118706537 @default.
- W3204118220 cites W2296319761 @default.
- W3204118220 cites W2399994860 @default.
- W3204118220 cites W2962715412 @default.
- W3204118220 cites W2962754331 @default.
- W3204118220 cites W2964121744 @default.
- W3204118220 cites W3118608800 @default.
- W3204118220 cites W61948969 @default.
- W3204118220 doi "https://doi.org/10.1109/ijcnn52387.2021.9534199" @default.
- W3204118220 hasPublicationYear "2021" @default.
- W3204118220 type Work @default.
- W3204118220 sameAs 3204118220 @default.
- W3204118220 citedByCount "0" @default.
- W3204118220 crossrefType "proceedings-article" @default.
- W3204118220 hasAuthorship W3204118220A5051693199 @default.
- W3204118220 hasAuthorship W3204118220A5085686472 @default.
- W3204118220 hasBestOaLocation W32041182202 @default.
- W3204118220 hasConcept C111030470 @default.
- W3204118220 hasConcept C11413529 @default.
- W3204118220 hasConcept C126255220 @default.
- W3204118220 hasConcept C134306372 @default.
- W3204118220 hasConcept C154945302 @default.
- W3204118220 hasConcept C169903167 @default.
- W3204118220 hasConcept C177148314 @default.
- W3204118220 hasConcept C194273485 @default.
- W3204118220 hasConcept C199360897 @default.
- W3204118220 hasConcept C2776135515 @default.
- W3204118220 hasConcept C2777735758 @default.
- W3204118220 hasConcept C33923547 @default.
- W3204118220 hasConcept C41008148 @default.
- W3204118220 hasConcept C50644808 @default.
- W3204118220 hasConcept C51632099 @default.
- W3204118220 hasConcept C555944384 @default.
- W3204118220 hasConcept C76155785 @default.
- W3204118220 hasConceptScore W3204118220C111030470 @default.
- W3204118220 hasConceptScore W3204118220C11413529 @default.
- W3204118220 hasConceptScore W3204118220C126255220 @default.
- W3204118220 hasConceptScore W3204118220C134306372 @default.
- W3204118220 hasConceptScore W3204118220C154945302 @default.
- W3204118220 hasConceptScore W3204118220C169903167 @default.
- W3204118220 hasConceptScore W3204118220C177148314 @default.
- W3204118220 hasConceptScore W3204118220C194273485 @default.
- W3204118220 hasConceptScore W3204118220C199360897 @default.
- W3204118220 hasConceptScore W3204118220C2776135515 @default.
- W3204118220 hasConceptScore W3204118220C2777735758 @default.
- W3204118220 hasConceptScore W3204118220C33923547 @default.
- W3204118220 hasConceptScore W3204118220C41008148 @default.
- W3204118220 hasConceptScore W3204118220C50644808 @default.
- W3204118220 hasConceptScore W3204118220C51632099 @default.
- W3204118220 hasConceptScore W3204118220C555944384 @default.
- W3204118220 hasConceptScore W3204118220C76155785 @default.
- W3204118220 hasFunder F4320332180 @default.
- W3204118220 hasLocation W32041182201 @default.
- W3204118220 hasLocation W32041182202 @default.
- W3204118220 hasOpenAccess W3204118220 @default.
- W3204118220 hasPrimaryLocation W32041182201 @default.
- W3204118220 hasRelatedWork W10634675 @default.
- W3204118220 hasRelatedWork W11141431 @default.
- W3204118220 hasRelatedWork W1268192 @default.
- W3204118220 hasRelatedWork W4861985 @default.
- W3204118220 hasRelatedWork W4919037 @default.
- W3204118220 hasRelatedWork W8259573 @default.
- W3204118220 hasRelatedWork W8897682 @default.
- W3204118220 hasRelatedWork W9443576 @default.
- W3204118220 hasRelatedWork W12792411 @default.
- W3204118220 hasRelatedWork W8078895 @default.
- W3204118220 isParatext "false" @default.
- W3204118220 isRetracted "false" @default.
- W3204118220 magId "3204118220" @default.
- W3204118220 workType "article" @default.