FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3094996462> ?p ?o ?g. }

• W3094996462 abstract "This work attempts to interpret modern deep (convolutional) networks from the principles of rate reduction and (shift) invariant classification. We show that the basic iterative gradient ascent scheme for maximizing the rate reduction of learned features naturally leads to a deep network, one iteration per layer. The architectures, operators (linear or nonlinear), and parameters of the network are all explicitly constructed layer-by-layer in a forward propagation fashion. All components of this ``white box'' network have precise optimization, statistical, and geometric interpretation. Our preliminary experiments indicate that such a network can already learn a good discriminative deep representation without any back propagation training. Moreover, all linear operators of the so-derived network naturally become multi-channel convolutions when we enforce classification to be rigorously shift-invariant. The derivation also indicates that such a convolutional network is significantly more efficient to learn and construct in the spectral domain." @default.
• W3094996462 created "2020-11-09" @default.
• W3094996462 creator A5019610791 @default.
• W3094996462 creator A5043328370 @default.
• W3094996462 creator A5055470034 @default.
• W3094996462 creator A5058405216 @default.
• W3094996462 creator A5062841020 @default.
• W3094996462 creator A5064317084 @default.
• W3094996462 date "2021-05-04" @default.
• W3094996462 modified "2023-09-23" @default.
• W3094996462 title "Deep Networks from the Principle of Rate Reduction" @default.
• W3094996462 cites W1922123711 @default.
• W3094996462 cites W1980451075 @default.
• W3094996462 cites W1990545423 @default.
• W3094996462 cites W2046779692 @default.
• W3094996462 cites W2064675550 @default.
• W3094996462 cites W2072072671 @default.
• W3094996462 cites W2093922090 @default.
• W3094996462 cites W2100556411 @default.
• W3094996462 cites W2101234009 @default.
• W3094996462 cites W2118103795 @default.
• W3094996462 cites W2129812935 @default.
• W3094996462 cites W2145889472 @default.
• W3094996462 cites W2153374008 @default.
• W3094996462 cites W2164931791 @default.
• W3094996462 cites W2172294131 @default.
• W3094996462 cites W2172654076 @default.
• W3094996462 cites W2194775991 @default.
• W3094996462 cites W2531409750 @default.
• W3094996462 cites W2549139847 @default.
• W3094996462 cites W2551814208 @default.
• W3094996462 cites W2553303224 @default.
• W3094996462 cites W2581624817 @default.
• W3094996462 cites W2752693045 @default.
• W3094996462 cites W2773726006 @default.
• W3094996462 cites W2807007689 @default.
• W3094996462 cites W2908974255 @default.
• W3094996462 cites W2951886768 @default.
• W3094996462 cites W2952054889 @default.
• W3094996462 cites W2958629450 @default.
• W3094996462 cites W2962835968 @default.
• W3094996462 cites W2963197835 @default.
• W3094996462 cites W2963367891 @default.
• W3094996462 cites W2963424732 @default.
• W3094996462 cites W2963446712 @default.
• W3094996462 cites W2963487351 @default.
• W3094996462 cites W2963699903 @default.
• W3094996462 cites W2963703618 @default.
• W3094996462 cites W2963755523 @default.
• W3094996462 cites W2963775850 @default.
• W3094996462 cites W2966661 @default.
• W3094996462 cites W2969945254 @default.
• W3094996462 cites W2970275575 @default.
• W3094996462 cites W2995038946 @default.
• W3094996462 cites W3008654699 @default.
• W3094996462 cites W3099751318 @default.
• W3094996462 cites W3102431071 @default.
• W3094996462 cites W3103253955 @default.
• W3094996462 cites W3120740533 @default.
• W3094996462 cites W3137695714 @default.
• W3094996462 cites W56210758 @default.
• W3094996462 cites W639309485 @default.
• W3094996462 cites W3034384538 @default.
• W3094996462 hasPublicationYear "2021" @default.
• W3094996462 type Work @default.
• W3094996462 sameAs 3094996462 @default.
• W3094996462 citedByCount "6" @default.
• W3094996462 countsByYear W30949964622020 @default.
• W3094996462 countsByYear W30949964622021 @default.
• W3094996462 crossrefType "journal-article" @default.
• W3094996462 hasAuthorship W3094996462A5019610791 @default.
• W3094996462 hasAuthorship W3094996462A5043328370 @default.
• W3094996462 hasAuthorship W3094996462A5055470034 @default.
• W3094996462 hasAuthorship W3094996462A5058405216 @default.
• W3094996462 hasAuthorship W3094996462A5062841020 @default.
• W3094996462 hasAuthorship W3094996462A5064317084 @default.
• W3094996462 hasConcept C108583219 @default.
• W3094996462 hasConcept C111335779 @default.
• W3094996462 hasConcept C11413529 @default.
• W3094996462 hasConcept C121332964 @default.
• W3094996462 hasConcept C126255220 @default.
• W3094996462 hasConcept C153180895 @default.
• W3094996462 hasConcept C154945302 @default.
• W3094996462 hasConcept C158622935 @default.
• W3094996462 hasConcept C17744445 @default.
• W3094996462 hasConcept C190470478 @default.
• W3094996462 hasConcept C199539241 @default.
• W3094996462 hasConcept C2524010 @default.
• W3094996462 hasConcept C2776359362 @default.
• W3094996462 hasConcept C33923547 @default.
• W3094996462 hasConcept C37914503 @default.
• W3094996462 hasConcept C41008148 @default.
• W3094996462 hasConcept C62520636 @default.
• W3094996462 hasConcept C81363708 @default.
• W3094996462 hasConcept C94625758 @default.
• W3094996462 hasConcept C97931131 @default.
• W3094996462 hasConceptScore W3094996462C108583219 @default.
• W3094996462 hasConceptScore W3094996462C111335779 @default.
• W3094996462 hasConceptScore W3094996462C11413529 @default.
• W3094996462 hasConceptScore W3094996462C121332964 @default.

• next