FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3101970908> ?p ?o ?g. }

• W3101970908 abstract "Federated Learning (FL) is a distributed learning paradigm that scales on-device learning collaboratively and privately. Standard FL algorithms such as FedAvg are primarily geared towards smooth unconstrained settings. In this paper, we study the Federated Composite Optimization (FCO) problem, in which the loss function contains a non-smooth regularizer. Such problems arise naturally in FL applications that involve sparsity, low-rank, monotonicity, or more general constraints. We first show that straightforward extensions of primal algorithms such as FedAvg are not well-suited for FCO since they suffer from the of primal averaging, resulting in poor convergence. As a solution, we propose a new primal-dual algorithm, Federated Dual Averaging (FedDualAvg), which by employing a novel server dual averaging procedure circumvents the curse of primal averaging. Our theoretical analysis and empirical experiments demonstrate that FedDualAvg outperforms the other baselines." @default.
• W3101970908 created "2020-11-23" @default.
• W3101970908 creator A5018931712 @default.
• W3101970908 creator A5082314161 @default.
• W3101970908 creator A5087483482 @default.
• W3101970908 date "2020-11-17" @default.
• W3101970908 modified "2023-09-27" @default.
• W3101970908 title "Federated Composite Optimization" @default.
• W3101970908 cites W131378802 @default.
• W3101970908 cites W1503425191 @default.
• W3101970908 cites W1505731132 @default.
• W3101970908 cites W1571416372 @default.
• W3101970908 cites W1616857247 @default.
• W3101970908 cites W1627780903 @default.
• W3101970908 cites W1788809966 @default.
• W3101970908 cites W1967138577 @default.
• W3101970908 cites W1969387119 @default.
• W3101970908 cites W2016384870 @default.
• W3101970908 cites W2030161963 @default.
• W3101970908 cites W2033468335 @default.
• W3101970908 cites W2044212084 @default.
• W3101970908 cites W2045185094 @default.
• W3101970908 cites W2063403497 @default.
• W3101970908 cites W2066332749 @default.
• W3101970908 cites W2078831169 @default.
• W3101970908 cites W2089632738 @default.
• W3101970908 cites W2096199223 @default.
• W3101970908 cites W2103972604 @default.
• W3101970908 cites W2115706991 @default.
• W3101970908 cites W2116612304 @default.
• W3101970908 cites W2130062883 @default.
• W3101970908 cites W2136885855 @default.
• W3101970908 cites W2140451424 @default.
• W3101970908 cites W2144839971 @default.
• W3101970908 cites W2148825261 @default.
• W3101970908 cites W2151591509 @default.
• W3101970908 cites W2163525955 @default.
• W3101970908 cites W2163803752 @default.
• W3101970908 cites W2164301055 @default.
• W3101970908 cites W2166706236 @default.
• W3101970908 cites W2167732364 @default.
• W3101970908 cites W2169713291 @default.
• W3101970908 cites W2170262790 @default.
• W3101970908 cites W2171413647 @default.
• W3101970908 cites W2216483973 @default.
• W3101970908 cites W2243169878 @default.
• W3101970908 cites W2271600034 @default.
• W3101970908 cites W2535496140 @default.
• W3101970908 cites W2607045744 @default.
• W3101970908 cites W2611328865 @default.
• W3101970908 cites W2616657226 @default.
• W3101970908 cites W2888561381 @default.
• W3101970908 cites W2900120080 @default.
• W3101970908 cites W2900182564 @default.
• W3101970908 cites W2902113386 @default.
• W3101970908 cites W2944305377 @default.
• W3101970908 cites W2944542720 @default.
• W3101970908 cites W2946270031 @default.
• W3101970908 cites W2962688627 @default.
• W3101970908 cites W2962939792 @default.
• W3101970908 cites W2962952793 @default.
• W3101970908 cites W2963179579 @default.
• W3101970908 cites W2963189955 @default.
• W3101970908 cites W2963300197 @default.
• W3101970908 cites W2963326510 @default.
• W3101970908 cites W2963794891 @default.
• W3101970908 cites W2964066689 @default.
• W3101970908 cites W2964191394 @default.
• W3101970908 cites W2970020383 @default.
• W3101970908 cites W2976335444 @default.
• W3101970908 cites W2989670708 @default.
• W3101970908 cites W2994684563 @default.
• W3101970908 cites W2995022099 @default.
• W3101970908 cites W2995653155 @default.
• W3101970908 cites W3008187686 @default.
• W3101970908 cites W3012968339 @default.
• W3101970908 cites W3016889113 @default.
• W3101970908 cites W3026202352 @default.
• W3101970908 cites W3028382483 @default.
• W3101970908 cites W3034666238 @default.
• W3101970908 cites W3034741284 @default.
• W3101970908 cites W3035507578 @default.
• W3101970908 cites W3037674069 @default.
• W3101970908 cites W3038022836 @default.
• W3101970908 cites W3094793624 @default.
• W3101970908 cites W3099314130 @default.
• W3101970908 cites W3100393648 @default.
• W3101970908 cites W3102089081 @default.
• W3101970908 cites W3102310167 @default.
• W3101970908 cites W3102816259 @default.
• W3101970908 cites W3103741939 @default.
• W3101970908 cites W3105347258 @default.
• W3101970908 cites W3105393233 @default.
• W3101970908 cites W3106287588 @default.
• W3101970908 cites W3119953551 @default.
• W3101970908 cites W905619 @default.
• W3101970908 cites W3122584554 @default.
• W3101970908 hasPublicationYear "2020" @default.
• W3101970908 type Work @default.
• W3101970908 sameAs 3101970908 @default.

• next