FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2947373402> ?p ?o ?g. }

• W2947373402 abstract "Backpropagation is driving today's artificial neural networks (ANNs). However, despite extensive research, it remains unclear if the brain implements this algorithm. Among neuroscientists, reinforcement learning (RL) algorithms are often seen as a realistic alternative: neurons can randomly introduce change, and use unspecific feedback signals to observe their effect on the cost and thus approximate their gradient. However, the convergence rate of such learning scales poorly with the number of involved neurons. Here we propose a hybrid learning approach. Each neuron uses an RL-type strategy to learn how to approximate the gradients that backpropagation would provide. We provide proof that our approach converges to the true gradient for certain classes of networks. In both feedforward and convolutional networks, we empirically show that our approach learns to approximate the gradient, and can match or the performance of exact gradient-based learning. Learning feedback weights provides a biologically plausible mechanism of achieving good performance, without the need for precise, pre-specified learning rules." @default.
• W2947373402 created "2019-06-07" @default.
• W2947373402 creator A5018372746 @default.
• W2947373402 creator A5039635544 @default.
• W2947373402 creator A5056565028 @default.
• W2947373402 date "2019-06-03" @default.
• W2947373402 modified "2023-09-26" @default.
• W2947373402 title "Learning to solve the credit assignment problem" @default.
• W2947373402 cites W1498436455 @default.
• W2947373402 cites W1677182931 @default.
• W2947373402 cites W1777649940 @default.
• W2947373402 cites W1931823119 @default.
• W2947373402 cites W1994016657 @default.
• W2947373402 cites W2018435387 @default.
• W2947373402 cites W2025768430 @default.
• W2947373402 cites W20283819 @default.
• W2947373402 cites W2041176801 @default.
• W2947373402 cites W2062670775 @default.
• W2947373402 cites W2063046089 @default.
• W2947373402 cites W2106437793 @default.
• W2947373402 cites W2111406701 @default.
• W2947373402 cites W2112439305 @default.
• W2947373402 cites W2117539524 @default.
• W2947373402 cites W2119717200 @default.
• W2947373402 cites W2134842679 @default.
• W2947373402 cites W2145339207 @default.
• W2947373402 cites W2155051950 @default.
• W2947373402 cites W2159110831 @default.
• W2947373402 cites W2162160137 @default.
• W2947373402 cites W2172115975 @default.
• W2947373402 cites W2263490141 @default.
• W2947373402 cites W2516591743 @default.
• W2947373402 cites W2527798464 @default.
• W2947373402 cites W2529004582 @default.
• W2947373402 cites W2536957931 @default.
• W2947373402 cites W2614634292 @default.
• W2947373402 cites W2766447205 @default.
• W2947373402 cites W2784022806 @default.
• W2947373402 cites W2789634097 @default.
• W2947373402 cites W2791461106 @default.
• W2947373402 cites W2795851676 @default.
• W2947373402 cites W2891772748 @default.
• W2947373402 cites W2899171226 @default.
• W2947373402 cites W2903723132 @default.
• W2947373402 cites W2916359281 @default.
• W2947373402 cites W2919115771 @default.
• W2947373402 cites W2920121562 @default.
• W2947373402 cites W2935958002 @default.
• W2947373402 cites W2945174841 @default.
• W2947373402 cites W2952448932 @default.
• W2947373402 cites W2952518779 @default.
• W2947373402 cites W2953293408 @default.
• W2947373402 cites W2962723887 @default.
• W2947373402 cites W2962734330 @default.
• W2947373402 cites W2962897886 @default.
• W2947373402 cites W2963416336 @default.
• W2947373402 cites W2963703360 @default.
• W2947373402 cites W2964115671 @default.
• W2947373402 cites W2964121744 @default.
• W2947373402 cites W3001063045 @default.
• W2947373402 cites W3005798058 @default.
• W2947373402 cites W3105708035 @default.
• W2947373402 cites W3118608800 @default.
• W2947373402 cites W3213472335 @default.
• W2947373402 hasPublicationYear "2019" @default.
• W2947373402 type Work @default.
• W2947373402 sameAs 2947373402 @default.
• W2947373402 citedByCount "3" @default.
• W2947373402 countsByYear W29473734022020 @default.
• W2947373402 countsByYear W29473734022021 @default.
• W2947373402 crossrefType "posted-content" @default.
• W2947373402 hasAuthorship W2947373402A5018372746 @default.
• W2947373402 hasAuthorship W2947373402A5039635544 @default.
• W2947373402 hasAuthorship W2947373402A5056565028 @default.
• W2947373402 hasConcept C119857082 @default.
• W2947373402 hasConcept C127413603 @default.
• W2947373402 hasConcept C133731056 @default.
• W2947373402 hasConcept C153258448 @default.
• W2947373402 hasConcept C154945302 @default.
• W2947373402 hasConcept C155032097 @default.
• W2947373402 hasConcept C162324750 @default.
• W2947373402 hasConcept C2777303404 @default.
• W2947373402 hasConcept C38858127 @default.
• W2947373402 hasConcept C41008148 @default.
• W2947373402 hasConcept C50522688 @default.
• W2947373402 hasConcept C50644808 @default.
• W2947373402 hasConcept C97541855 @default.
• W2947373402 hasConceptScore W2947373402C119857082 @default.
• W2947373402 hasConceptScore W2947373402C127413603 @default.
• W2947373402 hasConceptScore W2947373402C133731056 @default.
• W2947373402 hasConceptScore W2947373402C153258448 @default.
• W2947373402 hasConceptScore W2947373402C154945302 @default.
• W2947373402 hasConceptScore W2947373402C155032097 @default.
• W2947373402 hasConceptScore W2947373402C162324750 @default.
• W2947373402 hasConceptScore W2947373402C2777303404 @default.
• W2947373402 hasConceptScore W2947373402C38858127 @default.
• W2947373402 hasConceptScore W2947373402C41008148 @default.
• W2947373402 hasConceptScore W2947373402C50522688 @default.
• W2947373402 hasConceptScore W2947373402C50644808 @default.
• W2947373402 hasConceptScore W2947373402C97541855 @default.

• next