FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2893995718> ?p ?o ?g. }

• W2893995718 abstract "Despite advances in scalable models, the inference tools used for Gaussian processes (GPs) have yet to fully capitalize on developments in computing hardware. We present an efficient and general approach to GP inference based on Blackbox Matrix-Matrix multiplication (BBMM). BBMM inference uses a modified batched version of the conjugate gradients algorithm to derive all terms for training and inference in a single call. BBMM reduces the asymptotic complexity of exact GP inference from $O(n^3)$ to $O(n^2)$. Adapting this algorithm to scalable approximations and complex GP models simply requires a routine for efficient matrix-matrix multiplication with the kernel and its derivative. In addition, BBMM uses a specialized preconditioner to substantially speed up convergence. In experiments we show that BBMM effectively uses GPU hardware to dramatically accelerate both exact GP inference and scalable approximations. Additionally, we provide GPyTorch, a software platform for scalable GP inference via BBMM, built on PyTorch." @default.
• W2893995718 created "2018-10-05" @default.
• W2893995718 creator A5006380227 @default.
• W2893995718 creator A5043360924 @default.
• W2893995718 creator A5065203625 @default.
• W2893995718 creator A5072585411 @default.
• W2893995718 creator A5073697071 @default.
• W2893995718 date "2018-09-28" @default.
• W2893995718 modified "2023-09-24" @default.
• W2893995718 title "GPyTorch: Blackbox Matrix-Matrix Gaussian Process Inference with GPU Acceleration." @default.
• W2893995718 cites W137285897 @default.
• W2893995718 cites W1481900670 @default.
• W2893995718 cites W1571870753 @default.
• W2893995718 cites W1572063013 @default.
• W2893995718 cites W1746819321 @default.
• W2893995718 cites W1959608418 @default.
• W2893995718 cites W1997321933 @default.
• W2893995718 cites W2040728226 @default.
• W2893995718 cites W2099768828 @default.
• W2893995718 cites W2117686912 @default.
• W2893995718 cites W2119595900 @default.
• W2893995718 cites W2137135057 @default.
• W2893995718 cites W2170952786 @default.
• W2893995718 cites W2186615578 @default.
• W2893995718 cites W2194775991 @default.
• W2893995718 cites W2506995846 @default.
• W2893995718 cites W2523060838 @default.
• W2893995718 cites W2751238863 @default.
• W2893995718 cites W2764280570 @default.
• W2893995718 cites W2793399158 @default.
• W2893995718 cites W2899771611 @default.
• W2893995718 cites W2912811302 @default.
• W2893995718 cites W2949117887 @default.
• W2893995718 cites W2949568260 @default.
• W2893995718 cites W2950094539 @default.
• W2893995718 cites W2952291055 @default.
• W2893995718 cites W2952677397 @default.
• W2893995718 cites W2953234161 @default.
• W2893995718 cites W2963173418 @default.
• W2893995718 cites W2963395419 @default.
• W2893995718 cites W2963942108 @default.
• W2893995718 cites W2964135722 @default.
• W2893995718 cites W603018635 @default.
• W2893995718 hasPublicationYear "2018" @default.
• W2893995718 type Work @default.
• W2893995718 sameAs 2893995718 @default.
• W2893995718 citedByCount "97" @default.
• W2893995718 countsByYear W28939957182018 @default.
• W2893995718 countsByYear W28939957182019 @default.
• W2893995718 countsByYear W28939957182020 @default.
• W2893995718 countsByYear W28939957182021 @default.
• W2893995718 countsByYear W28939957182022 @default.
• W2893995718 crossrefType "posted-content" @default.
• W2893995718 hasAuthorship W2893995718A5006380227 @default.
• W2893995718 hasAuthorship W2893995718A5043360924 @default.
• W2893995718 hasAuthorship W2893995718A5065203625 @default.
• W2893995718 hasAuthorship W2893995718A5072585411 @default.
• W2893995718 hasAuthorship W2893995718A5073697071 @default.
• W2893995718 hasConcept C106487976 @default.
• W2893995718 hasConcept C11413529 @default.
• W2893995718 hasConcept C118615104 @default.
• W2893995718 hasConcept C121332964 @default.
• W2893995718 hasConcept C121684516 @default.
• W2893995718 hasConcept C154945302 @default.
• W2893995718 hasConcept C159694833 @default.
• W2893995718 hasConcept C159985019 @default.
• W2893995718 hasConcept C162324750 @default.
• W2893995718 hasConcept C163716315 @default.
• W2893995718 hasConcept C167431342 @default.
• W2893995718 hasConcept C17349429 @default.
• W2893995718 hasConcept C173608175 @default.
• W2893995718 hasConcept C192562407 @default.
• W2893995718 hasConcept C21442007 @default.
• W2893995718 hasConcept C2776214188 @default.
• W2893995718 hasConcept C2777303404 @default.
• W2893995718 hasConcept C33923547 @default.
• W2893995718 hasConcept C41008148 @default.
• W2893995718 hasConcept C459310 @default.
• W2893995718 hasConcept C48044578 @default.
• W2893995718 hasConcept C50522688 @default.
• W2893995718 hasConcept C50630238 @default.
• W2893995718 hasConcept C61326573 @default.
• W2893995718 hasConcept C62520636 @default.
• W2893995718 hasConcept C68339613 @default.
• W2893995718 hasConcept C74193536 @default.
• W2893995718 hasConcept C77088390 @default.
• W2893995718 hasConcept C80444323 @default.
• W2893995718 hasConcept C84114770 @default.
• W2893995718 hasConceptScore W2893995718C106487976 @default.
• W2893995718 hasConceptScore W2893995718C11413529 @default.
• W2893995718 hasConceptScore W2893995718C118615104 @default.
• W2893995718 hasConceptScore W2893995718C121332964 @default.
• W2893995718 hasConceptScore W2893995718C121684516 @default.
• W2893995718 hasConceptScore W2893995718C154945302 @default.
• W2893995718 hasConceptScore W2893995718C159694833 @default.
• W2893995718 hasConceptScore W2893995718C159985019 @default.
• W2893995718 hasConceptScore W2893995718C162324750 @default.
• W2893995718 hasConceptScore W2893995718C163716315 @default.
• W2893995718 hasConceptScore W2893995718C167431342 @default.
• W2893995718 hasConceptScore W2893995718C17349429 @default.

• next