Matches in SemOpenAlex for { <https://semopenalex.org/work/W4287029600> ?p ?o ?g. }
Showing items 1 to 83 of
83
with 100 items per page.
- W4287029600 abstract "Tensor models play an increasingly prominent role in many fields, notably in machine learning. In several applications, such as community detection, topic modeling and Gaussian mixture learning, one must estimate a low-rank signal from a noisy tensor. Hence, understanding the fundamental limits of estimators of that signal inevitably calls for the study of random tensors. Substantial progress has been recently achieved on this subject in the large-dimensional limit. Yet, some of the most significant among these results--in particular, a precise characterization of the abrupt phase transition (with respect to signal-to-noise ratio) that governs the performance of the maximum likelihood (ML) estimator of a symmetric rank-one model with Gaussian noise--were derived based of mean-field spin glass theory, which is not easily accessible to non-experts. In this work, we develop a sharply distinct and more elementary approach, relying on standard but powerful tools brought by years of advances in random matrix theory. The key idea is to study the spectra of random matrices arising from contractions of a given random tensor. We show how this gives access to spectral properties of the random tensor itself. For the aforementioned rank-one model, our technique yields a hitherto unknown fixed-point equation whose solution precisely matches the asymptotic performance of the ML estimator above the phase transition threshold in the third-order case. A numerical verification provides evidence that the same holds for orders 4 and 5, leading us to conjecture that, for any order, our fixed-point equation is equivalent to the known characterization of the ML estimation performance that had been obtained by relying on spin glasses. Moreover, our approach sheds light on certain properties of the ML problem landscape in large dimensions and can be extended to other models, such as asymmetric and non-Gaussian." @default.
- W4287029600 created "2022-07-25" @default.
- W4287029600 creator A5021257434 @default.
- W4287029600 creator A5033972557 @default.
- W4287029600 creator A5062382303 @default.
- W4287029600 date "2021-08-02" @default.
- W4287029600 modified "2023-10-16" @default.
- W4287029600 title "A Random Matrix Perspective on Random Tensors" @default.
- W4287029600 doi "https://doi.org/10.48550/arxiv.2108.00774" @default.
- W4287029600 hasPublicationYear "2021" @default.
- W4287029600 type Work @default.
- W4287029600 citedByCount "0" @default.
- W4287029600 crossrefType "posted-content" @default.
- W4287029600 hasAuthorship W4287029600A5021257434 @default.
- W4287029600 hasAuthorship W4287029600A5033972557 @default.
- W4287029600 hasAuthorship W4287029600A5062382303 @default.
- W4287029600 hasBestOaLocation W42870296001 @default.
- W4287029600 hasConcept C105795698 @default.
- W4287029600 hasConcept C106487976 @default.
- W4287029600 hasConcept C114614502 @default.
- W4287029600 hasConcept C115961682 @default.
- W4287029600 hasConcept C121332964 @default.
- W4287029600 hasConcept C121864883 @default.
- W4287029600 hasConcept C130402806 @default.
- W4287029600 hasConcept C134306372 @default.
- W4287029600 hasConcept C151201525 @default.
- W4287029600 hasConcept C154945302 @default.
- W4287029600 hasConcept C155281189 @default.
- W4287029600 hasConcept C158693339 @default.
- W4287029600 hasConcept C159985019 @default.
- W4287029600 hasConcept C163716315 @default.
- W4287029600 hasConcept C164226766 @default.
- W4287029600 hasConcept C166785042 @default.
- W4287029600 hasConcept C185429906 @default.
- W4287029600 hasConcept C192562407 @default.
- W4287029600 hasConcept C202444582 @default.
- W4287029600 hasConcept C28826006 @default.
- W4287029600 hasConcept C33923547 @default.
- W4287029600 hasConcept C41008148 @default.
- W4287029600 hasConcept C62520636 @default.
- W4287029600 hasConcept C64812099 @default.
- W4287029600 hasConcept C99498987 @default.
- W4287029600 hasConceptScore W4287029600C105795698 @default.
- W4287029600 hasConceptScore W4287029600C106487976 @default.
- W4287029600 hasConceptScore W4287029600C114614502 @default.
- W4287029600 hasConceptScore W4287029600C115961682 @default.
- W4287029600 hasConceptScore W4287029600C121332964 @default.
- W4287029600 hasConceptScore W4287029600C121864883 @default.
- W4287029600 hasConceptScore W4287029600C130402806 @default.
- W4287029600 hasConceptScore W4287029600C134306372 @default.
- W4287029600 hasConceptScore W4287029600C151201525 @default.
- W4287029600 hasConceptScore W4287029600C154945302 @default.
- W4287029600 hasConceptScore W4287029600C155281189 @default.
- W4287029600 hasConceptScore W4287029600C158693339 @default.
- W4287029600 hasConceptScore W4287029600C159985019 @default.
- W4287029600 hasConceptScore W4287029600C163716315 @default.
- W4287029600 hasConceptScore W4287029600C164226766 @default.
- W4287029600 hasConceptScore W4287029600C166785042 @default.
- W4287029600 hasConceptScore W4287029600C185429906 @default.
- W4287029600 hasConceptScore W4287029600C192562407 @default.
- W4287029600 hasConceptScore W4287029600C202444582 @default.
- W4287029600 hasConceptScore W4287029600C28826006 @default.
- W4287029600 hasConceptScore W4287029600C33923547 @default.
- W4287029600 hasConceptScore W4287029600C41008148 @default.
- W4287029600 hasConceptScore W4287029600C62520636 @default.
- W4287029600 hasConceptScore W4287029600C64812099 @default.
- W4287029600 hasConceptScore W4287029600C99498987 @default.
- W4287029600 hasLocation W42870296001 @default.
- W4287029600 hasOpenAccess W4287029600 @default.
- W4287029600 hasPrimaryLocation W42870296001 @default.
- W4287029600 hasRelatedWork W1980746620 @default.
- W4287029600 hasRelatedWork W2042585404 @default.
- W4287029600 hasRelatedWork W2154584674 @default.
- W4287029600 hasRelatedWork W2546128363 @default.
- W4287029600 hasRelatedWork W2566121622 @default.
- W4287029600 hasRelatedWork W2896055607 @default.
- W4287029600 hasRelatedWork W3004303341 @default.
- W4287029600 hasRelatedWork W3009576629 @default.
- W4287029600 hasRelatedWork W4289375656 @default.
- W4287029600 hasRelatedWork W4307626464 @default.
- W4287029600 isParatext "false" @default.
- W4287029600 isRetracted "false" @default.
- W4287029600 workType "article" @default.