FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3120949090> ?p ?o ?g. }

• W3120949090 abstract "We propose reproducing activation functions (RAFs) to improve deep learning accuracy for various applications ranging from computer vision to scientific computing. The idea is to employ several basic functions and their learnable linear combination to construct neuron-wise data-driven activation functions for each neuron. Armed with RAFs, neural networks (NNs) can reproduce traditional approximation tools and, therefore, approximate target functions with a smaller number of parameters than traditional NNs. In NN training, RAFs can generate neural tangent kernels (NTKs) with a better condition number than traditional activation functions lessening the spectral bias of deep learning. As demonstrated by extensive numerical tests, the proposed RAFs can facilitate the convergence of deep learning optimization for a solution with higher accuracy than existing deep learning solvers for audio/image/video reconstruction, PDEs, and eigenvalue problems. With RAFs, the errors of audio/video reconstruction, PDEs, and eigenvalue problems are decreased by over 14%, 73%, 99%, respectively, compared with baseline, while the performance of image reconstruction increases by 58%." @default.
• W3120949090 created "2021-01-18" @default.
• W3120949090 creator A5004525330 @default.
• W3120949090 creator A5075861619 @default.
• W3120949090 creator A5079602544 @default.
• W3120949090 creator A5081680496 @default.
• W3120949090 date "2021-01-12" @default.
• W3120949090 modified "2023-09-27" @default.
• W3120949090 title "Reproducing Activation Function for Deep Learning" @default.
• W3120949090 cites W1522301498 @default.
• W3120949090 cites W2146502635 @default.
• W3120949090 cites W2151029520 @default.
• W3120949090 cites W2166116275 @default.
• W3120949090 cites W2194775991 @default.
• W3120949090 cites W2251586653 @default.
• W3120949090 cites W2528305538 @default.
• W3120949090 cites W253307310 @default.
• W3120949090 cites W2533570097 @default.
• W3120949090 cites W2550848904 @default.
• W3120949090 cites W2557283755 @default.
• W3120949090 cites W2613715972 @default.
• W3120949090 cites W2760972773 @default.
• W3120949090 cites W2780503584 @default.
• W3120949090 cites W2803629276 @default.
• W3120949090 cites W2809090039 @default.
• W3120949090 cites W2810756260 @default.
• W3120949090 cites W2894604724 @default.
• W3120949090 cites W2899253307 @default.
• W3120949090 cites W2899283552 @default.
• W3120949090 cites W2899507754 @default.
• W3120949090 cites W2899748887 @default.
• W3120949090 cites W2902123811 @default.
• W3120949090 cites W2911867426 @default.
• W3120949090 cites W2920266844 @default.
• W3120949090 cites W2923130667 @default.
• W3120949090 cites W2944736082 @default.
• W3120949090 cites W2948551291 @default.
• W3120949090 cites W2951651062 @default.
• W3120949090 cites W2962849139 @default.
• W3120949090 cites W2963220074 @default.
• W3120949090 cites W2963627347 @default.
• W3120949090 cites W2970217468 @default.
• W3120949090 cites W2971278627 @default.
• W3120949090 cites W2978175238 @default.
• W3120949090 cites W2981978060 @default.
• W3120949090 cites W2985622559 @default.
• W3120949090 cites W2988621382 @default.
• W3120949090 cites W2989630530 @default.
• W3120949090 cites W2993505246 @default.
• W3120949090 cites W2996338702 @default.
• W3120949090 cites W3000236721 @default.
• W3120949090 cites W3000403725 @default.
• W3120949090 cites W3004547276 @default.
• W3120949090 cites W3011537551 @default.
• W3120949090 cites W3014573765 @default.
• W3120949090 cites W3014618972 @default.
• W3120949090 cites W3023241366 @default.
• W3120949090 cites W3031010353 @default.
• W3120949090 cites W3034259269 @default.
• W3120949090 cites W3035591705 @default.
• W3120949090 cites W3036843665 @default.
• W3120949090 cites W3037386592 @default.
• W3120949090 cites W3041183578 @default.
• W3120949090 cites W3046173836 @default.
• W3120949090 cites W3047103993 @default.
• W3120949090 cites W3093990252 @default.
• W3120949090 cites W3097724114 @default.
• W3120949090 cites W3099710307 @default.
• W3120949090 cites W3103313582 @default.
• W3120949090 cites W3104183394 @default.
• W3120949090 cites W3107384982 @default.
• W3120949090 cites W3161792008 @default.
• W3120949090 cites W3191617243 @default.
• W3120949090 cites W3197818647 @default.
• W3120949090 cites W2991401328 @default.
• W3120949090 doi "https://doi.org/10.48550/arxiv.2101.04844" @default.
• W3120949090 hasPublicationYear "2021" @default.
• W3120949090 type Work @default.
• W3120949090 sameAs 3120949090 @default.
• W3120949090 citedByCount "5" @default.
• W3120949090 countsByYear W31209490902021 @default.
• W3120949090 crossrefType "posted-content" @default.
• W3120949090 hasAuthorship W3120949090A5004525330 @default.
• W3120949090 hasAuthorship W3120949090A5075861619 @default.
• W3120949090 hasAuthorship W3120949090A5079602544 @default.
• W3120949090 hasAuthorship W3120949090A5081680496 @default.
• W3120949090 hasBestOaLocation W31209490901 @default.
• W3120949090 hasConcept C108583219 @default.
• W3120949090 hasConcept C11413529 @default.
• W3120949090 hasConcept C115051666 @default.
• W3120949090 hasConcept C121332964 @default.
• W3120949090 hasConcept C14036430 @default.
• W3120949090 hasConcept C153180895 @default.
• W3120949090 hasConcept C154945302 @default.
• W3120949090 hasConcept C158693339 @default.
• W3120949090 hasConcept C162324750 @default.
• W3120949090 hasConcept C2777303404 @default.
• W3120949090 hasConcept C2984842247 @default.
• W3120949090 hasConcept C38365724 @default.
• W3120949090 hasConcept C41008148 @default.

• next