Matches in SemOpenAlex for { <https://semopenalex.org/work/W3036772641> ?p ?o ?g. }
- W3036772641 abstract "Kernel methods provide an elegant and principled approach to nonparametric learning, but so far could hardly be used in large scale problems, since naive implementations scale poorly with data size. Recent advances have shown the benefits of a number of algorithmic ideas, for example combining optimization, numerical linear algebra and random projections. Here, we push these efforts further to develop and test a solver that takes full advantage of GPU hardware. Towards this end, we designed a preconditioned gradient solver for kernel methods exploiting both GPU acceleration and parallelization with multiple GPUs, implementing out-of-core variants of common linear algebra operations to guarantee optimal hardware utilization. Further, we optimize the numerical precision of different operations and maximize efficiency of matrix-vector multiplications. As a result we can experimentally show dramatic speedups on datasets with billions of points, while still guaranteeing state of the art performance. Additionally, we make our software available as an easy to use library." @default.
- W3036772641 created "2020-06-25" @default.
- W3036772641 creator A5013505114 @default.
- W3036772641 creator A5014399020 @default.
- W3036772641 creator A5019263367 @default.
- W3036772641 creator A5061220999 @default.
- W3036772641 date "2020-06-18" @default.
- W3036772641 modified "2023-09-24" @default.
- W3036772641 title "Kernel methods through the roof: handling billions of points efficiently" @default.
- W3036772641 cites W1510073064 @default.
- W3036772641 cites W1533660737 @default.
- W3036772641 cites W1540155273 @default.
- W3036772641 cites W1551209770 @default.
- W3036772641 cites W1558138778 @default.
- W3036772641 cites W1560724230 @default.
- W3036772641 cites W1698155719 @default.
- W3036772641 cites W1702387805 @default.
- W3036772641 cites W1746819321 @default.
- W3036772641 cites W1777124189 @default.
- W3036772641 cites W194417083 @default.
- W3036772641 cites W1992774275 @default.
- W3036772641 cites W2002355073 @default.
- W3036772641 cites W2012501405 @default.
- W3036772641 cites W2047278710 @default.
- W3036772641 cites W2107791152 @default.
- W3036772641 cites W2112545207 @default.
- W3036772641 cites W2118563516 @default.
- W3036772641 cites W2123395972 @default.
- W3036772641 cites W2141566892 @default.
- W3036772641 cites W2144902422 @default.
- W3036772641 cites W2145763471 @default.
- W3036772641 cites W2159350554 @default.
- W3036772641 cites W2160840682 @default.
- W3036772641 cites W2162525251 @default.
- W3036772641 cites W2278578264 @default.
- W3036772641 cites W2549019841 @default.
- W3036772641 cites W2550412943 @default.
- W3036772641 cites W2568875900 @default.
- W3036772641 cites W2616038372 @default.
- W3036772641 cites W2728139190 @default.
- W3036772641 cites W2782213427 @default.
- W3036772641 cites W2791746604 @default.
- W3036772641 cites W2809090039 @default.
- W3036772641 cites W2891234027 @default.
- W3036772641 cites W2891583539 @default.
- W3036772641 cites W2913475682 @default.
- W3036772641 cites W2953256123 @default.
- W3036772641 cites W2961262728 @default.
- W3036772641 cites W2963013450 @default.
- W3036772641 cites W2963190151 @default.
- W3036772641 cites W2963395419 @default.
- W3036772641 cites W2963518317 @default.
- W3036772641 cites W2963662217 @default.
- W3036772641 cites W2963674301 @default.
- W3036772641 cites W2963711523 @default.
- W3036772641 cites W2963756295 @default.
- W3036772641 cites W2963899745 @default.
- W3036772641 cites W2963942108 @default.
- W3036772641 cites W2963972841 @default.
- W3036772641 cites W2964220529 @default.
- W3036772641 cites W2970455294 @default.
- W3036772641 cites W2970971581 @default.
- W3036772641 cites W2971353969 @default.
- W3036772641 cites W3010435838 @default.
- W3036772641 cites W3037025065 @default.
- W3036772641 cites W579519726 @default.
- W3036772641 cites W607505555 @default.
- W3036772641 hasPublicationYear "2020" @default.
- W3036772641 type Work @default.
- W3036772641 sameAs 3036772641 @default.
- W3036772641 citedByCount "4" @default.
- W3036772641 countsByYear W30367726412021 @default.
- W3036772641 countsByYear W30367726412022 @default.
- W3036772641 crossrefType "posted-content" @default.
- W3036772641 hasAuthorship W3036772641A5013505114 @default.
- W3036772641 hasAuthorship W3036772641A5014399020 @default.
- W3036772641 hasAuthorship W3036772641A5019263367 @default.
- W3036772641 hasAuthorship W3036772641A5061220999 @default.
- W3036772641 hasConcept C11413529 @default.
- W3036772641 hasConcept C114614502 @default.
- W3036772641 hasConcept C134306372 @default.
- W3036772641 hasConcept C139352143 @default.
- W3036772641 hasConcept C163834973 @default.
- W3036772641 hasConcept C173608175 @default.
- W3036772641 hasConcept C199360897 @default.
- W3036772641 hasConcept C2524010 @default.
- W3036772641 hasConcept C26713055 @default.
- W3036772641 hasConcept C2778119891 @default.
- W3036772641 hasConcept C2778770139 @default.
- W3036772641 hasConcept C33923547 @default.
- W3036772641 hasConcept C41008148 @default.
- W3036772641 hasConcept C459310 @default.
- W3036772641 hasConcept C6802819 @default.
- W3036772641 hasConcept C74193536 @default.
- W3036772641 hasConcept C80444323 @default.
- W3036772641 hasConceptScore W3036772641C11413529 @default.
- W3036772641 hasConceptScore W3036772641C114614502 @default.
- W3036772641 hasConceptScore W3036772641C134306372 @default.
- W3036772641 hasConceptScore W3036772641C139352143 @default.
- W3036772641 hasConceptScore W3036772641C163834973 @default.