FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2849120009> ?p ?o ?g. }

• W2849120009 abstract "The problem of distributed representation learning is one in which multiple sources of information $X_1,ldots,X_K$ are processed separately so as to learn as much information as possible about some ground truth $Y$. We investigate this problem from information-theoretic grounds, through a generalization of Tishby's centralized Information Bottleneck (IB) method to the distributed setting. Specifically, $K$ encoders, $K geq 2$, compress their observations $X_1,ldots,X_K$ separately in a manner such that, collectively, the produced representations preserve as much information as possible about $Y$. We study both discrete memoryless (DM) and memoryless vector Gaussian data models. For the discrete model, we establish a single-letter characterization of the optimal tradeoff between complexity (or rate) and relevance (or information) for a class of memoryless sources (the observations $X_1,ldots,X_K$ being conditionally independent given $Y$). For the vector Gaussian model, we provide an explicit characterization of the optimal complexity-relevance tradeoff. Furthermore, we develop a variational bound on the complexity-relevance tradeoff which generalizes the evidence lower bound (ELBO) to the distributed setting. We also provide two algorithms that allow to compute this bound: i) a Blahut-Arimoto type iterative algorithm which enables to compute optimal complexity-relevance encoding mappings by iterating over a set of self-consistent equations, and ii) a variational inference type algorithm in which the encoding mappings are parametrized by neural networks and the bound approximated by Markov sampling and optimized with stochastic gradient descent. Numerical results on synthetic and real datasets are provided to support the efficiency of the approaches and algorithms developed in this paper." @default.
• W2849120009 created "2018-07-19" @default.
• W2849120009 creator A5014817346 @default.
• W2849120009 creator A5014906072 @default.
• W2849120009 date "2018-07-11" @default.
• W2849120009 modified "2023-09-27" @default.
• W2849120009 title "Distributed Variational Representation Learning" @default.
• W2849120009 cites W1494099594 @default.
• W2849120009 cites W1533861849 @default.
• W2849120009 cites W1570963478 @default.
• W2849120009 cites W1622633983 @default.
• W2849120009 cites W1670132599 @default.
• W2849120009 cites W1874027545 @default.
• W2849120009 cites W1959608418 @default.
• W2849120009 cites W1966168239 @default.
• W2849120009 cites W1989151402 @default.
• W2849120009 cites W1990293992 @default.
• W2849120009 cites W2047229728 @default.
• W2849120009 cites W2048679005 @default.
• W2849120009 cites W2049633694 @default.
• W2849120009 cites W2050968963 @default.
• W2849120009 cites W2056129277 @default.
• W2849120009 cites W2060314721 @default.
• W2849120009 cites W206162032 @default.
• W2849120009 cites W2070945723 @default.
• W2849120009 cites W2079517420 @default.
• W2849120009 cites W2099111195 @default.
• W2849120009 cites W2101324110 @default.
• W2849120009 cites W2109743529 @default.
• W2849120009 cites W2114085948 @default.
• W2849120009 cites W2114771311 @default.
• W2849120009 cites W2122759946 @default.
• W2849120009 cites W2122925692 @default.
• W2849120009 cites W2129625650 @default.
• W2849120009 cites W2136288990 @default.
• W2849120009 cites W2139338362 @default.
• W2849120009 cites W2143121893 @default.
• W2849120009 cites W2148986322 @default.
• W2849120009 cites W2156872152 @default.
• W2849120009 cites W2156875677 @default.
• W2849120009 cites W2156909104 @default.
• W2849120009 cites W2163166770 @default.
• W2849120009 cites W2163922914 @default.
• W2849120009 cites W2170503197 @default.
• W2849120009 cites W2173267959 @default.
• W2849120009 cites W2179036435 @default.
• W2849120009 cites W2254691124 @default.
• W2849120009 cites W2398870399 @default.
• W2849120009 cites W2431962807 @default.
• W2849120009 cites W2527691228 @default.
• W2849120009 cites W2547875792 @default.
• W2849120009 cites W2548228487 @default.
• W2849120009 cites W2557579533 @default.
• W2849120009 cites W2582181367 @default.
• W2849120009 cites W2683470288 @default.
• W2849120009 cites W27434444 @default.
• W2849120009 cites W2753738274 @default.
• W2849120009 cites W2785885194 @default.
• W2849120009 cites W2788304628 @default.
• W2849120009 cites W2788440360 @default.
• W2849120009 cites W2789776893 @default.
• W2849120009 cites W2797621078 @default.
• W2849120009 cites W2809994596 @default.
• W2849120009 cites W282155574 @default.
• W2849120009 cites W2828552710 @default.
• W2849120009 cites W2962761397 @default.
• W2849120009 cites W2962897886 @default.
• W2849120009 cites W2963047245 @default.
• W2849120009 cites W2963386266 @default.
• W2849120009 cites W2963743310 @default.
• W2849120009 cites W2963862692 @default.
• W2849120009 cites W2964121744 @default.
• W2849120009 cites W2964209830 @default.
• W2849120009 cites W2979454998 @default.
• W2849120009 cites W2982267762 @default.
• W2849120009 cites W2996935267 @default.
• W2849120009 cites W3098445100 @default.
• W2849120009 cites W3099602767 @default.
• W2849120009 cites W565868953 @default.
• W2849120009 hasPublicationYear "2018" @default.
• W2849120009 type Work @default.
• W2849120009 sameAs 2849120009 @default.
• W2849120009 citedByCount "1" @default.
• W2849120009 countsByYear W28491200092017 @default.
• W2849120009 crossrefType "posted-content" @default.
• W2849120009 hasAuthorship W2849120009A5014817346 @default.
• W2849120009 hasAuthorship W2849120009A5014906072 @default.
• W2849120009 hasConcept C11413529 @default.
• W2849120009 hasConcept C118615104 @default.
• W2849120009 hasConcept C121332964 @default.
• W2849120009 hasConcept C134306372 @default.
• W2849120009 hasConcept C152139883 @default.
• W2849120009 hasConcept C154945302 @default.
• W2849120009 hasConcept C158154518 @default.
• W2849120009 hasConcept C163716315 @default.
• W2849120009 hasConcept C177148314 @default.
• W2849120009 hasConcept C17744445 @default.
• W2849120009 hasConcept C179799912 @default.
• W2849120009 hasConcept C199539241 @default.
• W2849120009 hasConcept C2776359362 @default.

• next