FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3007969445> ?p ?o ?g. }

• W3007969445 abstract "Graph representation learning is a ubiquitous task in machine learning where the goal is to embed each vertex into a low-dimensional vector space. We consider the bipartite graph and formalize its representation learning problem as a statistical estimation problem of parameters in a semiparametric exponential family distribution. The bipartite graph is assumed to be generated by a semiparametric exponential family distribution, whose parametric component is given by the proximity of outputs of two one-layer neural networks, while nonparametric (nuisance) component is the base measure. Neural networks take high-dimensional features as inputs and output embedding vectors. In this setting, the representation learning problem is equivalent to recovering the weight matrices. The main challenges of estimation arise from the nonlinearity of activation functions and the nonparametric nuisance component of the distribution. To overcome these challenges, we propose a pseudo-likelihood objective based on the rank-order decomposition technique and focus on its local geometry. We show that the proposed objective is strongly convex in a neighborhood around the ground truth, so that a gradient descent-based method achieves linear convergence rate. Moreover, we prove that the sample complexity of the problem is linear in dimensions (up to logarithmic factors), which is consistent with parametric Gaussian models. However, our estimator is robust to any model misspecification within the exponential family, which is validated in extensive experiments." @default.
• W3007969445 created "2020-03-06" @default.
• W3007969445 creator A5008573881 @default.
• W3007969445 creator A5016868056 @default.
• W3007969445 creator A5048272675 @default.
• W3007969445 creator A5056179845 @default.
• W3007969445 creator A5078210646 @default.
• W3007969445 date "2020-03-02" @default.
• W3007969445 modified "2023-10-16" @default.
• W3007969445 title "Semiparametric Nonlinear Bipartite Graph Representation Learning with Provable Guarantees" @default.
• W3007969445 cites W136217777 @default.
• W3007969445 cites W1488211288 @default.
• W3007969445 cites W1495663238 @default.
• W3007969445 cites W1538151859 @default.
• W3007969445 cites W1639316992 @default.
• W3007969445 cites W1873595945 @default.
• W3007969445 cites W188608978 @default.
• W3007969445 cites W1971788485 @default.
• W3007969445 cites W1971800148 @default.
• W3007969445 cites W2045901075 @default.
• W3007969445 cites W2052044664 @default.
• W3007969445 cites W2052819443 @default.
• W3007969445 cites W2094811636 @default.
• W3007969445 cites W2108753466 @default.
• W3007969445 cites W2116341502 @default.
• W3007969445 cites W2127345773 @default.
• W3007969445 cites W2142535891 @default.
• W3007969445 cites W2156718197 @default.
• W3007969445 cites W2157331557 @default.
• W3007969445 cites W2163922914 @default.
• W3007969445 cites W2242161203 @default.
• W3007969445 cites W2350220533 @default.
• W3007969445 cites W2401757692 @default.
• W3007969445 cites W2519716980 @default.
• W3007969445 cites W2612872092 @default.
• W3007969445 cites W2624407581 @default.
• W3007969445 cites W2626325961 @default.
• W3007969445 cites W2767849480 @default.
• W3007969445 cites W2790197675 @default.
• W3007969445 cites W2804694832 @default.
• W3007969445 cites W2886067286 @default.
• W3007969445 cites W2890653522 @default.
• W3007969445 cites W2893671662 @default.
• W3007969445 cites W2902028816 @default.
• W3007969445 cites W2904630684 @default.
• W3007969445 cites W2905224888 @default.
• W3007969445 cites W2962714108 @default.
• W3007969445 cites W2962756421 @default.
• W3007969445 cites W2962767131 @default.
• W3007969445 cites W2962767366 @default.
• W3007969445 cites W2962797944 @default.
• W3007969445 cites W2963177480 @default.
• W3007969445 cites W2963380955 @default.
• W3007969445 cites W2963460103 @default.
• W3007969445 cites W2963744427 @default.
• W3007969445 cites W2963829083 @default.
• W3007969445 cites W2963881378 @default.
• W3007969445 cites W2964135750 @default.
• W3007969445 cites W2964146881 @default.
• W3007969445 cites W2970330753 @default.
• W3007969445 cites W2970862128 @default.
• W3007969445 cites W2970889590 @default.
• W3007969445 cites W2972993921 @default.
• W3007969445 cites W3007611469 @default.
• W3007969445 cites W3099973750 @default.
• W3007969445 cites W3103995645 @default.
• W3007969445 cites W3104097132 @default.
• W3007969445 cites W3144088146 @default.
• W3007969445 cites W34646664 @default.
• W3007969445 cites W602552144 @default.
• W3007969445 hasPublicationYear "2020" @default.
• W3007969445 type Work @default.
• W3007969445 sameAs 3007969445 @default.
• W3007969445 citedByCount "0" @default.
• W3007969445 crossrefType "posted-content" @default.
• W3007969445 hasAuthorship W3007969445A5008573881 @default.
• W3007969445 hasAuthorship W3007969445A5016868056 @default.
• W3007969445 hasAuthorship W3007969445A5048272675 @default.
• W3007969445 hasAuthorship W3007969445A5056179845 @default.
• W3007969445 hasAuthorship W3007969445A5078210646 @default.
• W3007969445 hasConcept C105795698 @default.
• W3007969445 hasConcept C117251300 @default.
• W3007969445 hasConcept C126255220 @default.
• W3007969445 hasConcept C154945302 @default.
• W3007969445 hasConcept C185429906 @default.
• W3007969445 hasConcept C28826006 @default.
• W3007969445 hasConcept C33923547 @default.
• W3007969445 hasConcept C41008148 @default.
• W3007969445 hasConcept C55974624 @default.
• W3007969445 hasConcept C59404180 @default.
• W3007969445 hasConceptScore W3007969445C105795698 @default.
• W3007969445 hasConceptScore W3007969445C117251300 @default.
• W3007969445 hasConceptScore W3007969445C126255220 @default.
• W3007969445 hasConceptScore W3007969445C154945302 @default.
• W3007969445 hasConceptScore W3007969445C185429906 @default.
• W3007969445 hasConceptScore W3007969445C28826006 @default.
• W3007969445 hasConceptScore W3007969445C33923547 @default.
• W3007969445 hasConceptScore W3007969445C41008148 @default.
• W3007969445 hasConceptScore W3007969445C55974624 @default.
• W3007969445 hasConceptScore W3007969445C59404180 @default.

• next