FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3134581569> ?p ?o ?g. }

• W3134581569 abstract "A widely recognized difficulty in federated learning arises from the statistical heterogeneity among clients: local datasets often come from different but not entirely unrelated distributions, and personalization is, therefore, necessary to achieve optimal results from each individual's perspective. In this paper, we show how the excess risks of personalized federated learning with a smooth, strongly convex loss depend on data heterogeneity from a minimax point of view. Our analysis reveals a surprising theorem of the alternative for personalized federated learning: there exists a threshold such that (a) if a certain measure of data heterogeneity is below this threshold, the FedAvg algorithm [McMahan et al., 2017] is minimax optimal; (b) when the measure of heterogeneity is above this threshold, then doing pure local training (i.e., clients solve empirical risk minimization problems on their local datasets without any communication) is minimax optimal. As an implication, our results show that the presumably difficult (infinite-dimensional) problem of adapting to client-wise heterogeneity can be reduced to a simple binary decision problem of choosing between the two baseline algorithms. Our analysis relies on a new notion of algorithmic stability that takes into account the nature of federated learning." @default.
• W3134581569 created "2021-03-15" @default.
• W3134581569 creator A5000011776 @default.
• W3134581569 creator A5005455141 @default.
• W3134581569 creator A5068556649 @default.
• W3134581569 creator A5080575294 @default.
• W3134581569 date "2021-03-02" @default.
• W3134581569 modified "2023-09-27" @default.
• W3134581569 title "A Theorem of the Alternative for Personalized Federated Learning." @default.
• W3134581569 cites W1516903196 @default.
• W3134581569 cites W1538452572 @default.
• W3134581569 cites W1845277745 @default.
• W3134581569 cites W2040852548 @default.
• W3134581569 cites W2057681237 @default.
• W3134581569 cites W2096840748 @default.
• W3134581569 cites W2104094955 @default.
• W3134581569 cites W2112269233 @default.
• W3134581569 cites W2124231679 @default.
• W3134581569 cites W2131953535 @default.
• W3134581569 cites W2139338362 @default.
• W3134581569 cites W2143104527 @default.
• W3134581569 cites W2160257547 @default.
• W3134581569 cites W2165698076 @default.
• W3134581569 cites W2237537322 @default.
• W3134581569 cites W2328226763 @default.
• W3134581569 cites W2541884796 @default.
• W3134581569 cites W2604763608 @default.
• W3134581569 cites W2787717425 @default.
• W3134581569 cites W2803867449 @default.
• W3134581569 cites W2890870050 @default.
• W3134581569 cites W2913340405 @default.
• W3134581569 cites W2913743069 @default.
• W3134581569 cites W2914328083 @default.
• W3134581569 cites W2923966100 @default.
• W3134581569 cites W2925446857 @default.
• W3134581569 cites W2948257377 @default.
• W3134581569 cites W2965497096 @default.
• W3134581569 cites W2970814018 @default.
• W3134581569 cites W2972497108 @default.
• W3134581569 cites W2976335444 @default.
• W3134581569 cites W2980216952 @default.
• W3134581569 cites W2982654255 @default.
• W3134581569 cites W2990709181 @default.
• W3134581569 cites W2990789643 @default.
• W3134581569 cites W2995022099 @default.
• W3134581569 cites W2995653155 @default.
• W3134581569 cites W3005776401 @default.
• W3134581569 cites W3006157707 @default.
• W3134581569 cites W3006977545 @default.
• W3134581569 cites W3007345209 @default.
• W3134581569 cites W3007548213 @default.
• W3134581569 cites W3007684729 @default.
• W3134581569 cites W3012968339 @default.
• W3134581569 cites W3033462932 @default.
• W3134581569 cites W3034553394 @default.
• W3134581569 cites W3035353486 @default.
• W3134581569 cites W3035870018 @default.
• W3134581569 cites W3035936437 @default.
• W3134581569 cites W3035990676 @default.
• W3134581569 cites W3037123797 @default.
• W3134581569 cites W3037674069 @default.
• W3134581569 cites W3038028469 @default.
• W3134581569 cites W3046037042 @default.
• W3134581569 cites W3089578458 @default.
• W3134581569 cites W3092407976 @default.
• W3134581569 cites W3094219710 @default.
• W3134581569 cites W3097700865 @default.
• W3134581569 cites W3100393648 @default.
• W3134581569 cites W3102310167 @default.
• W3134581569 cites W3106287588 @default.
• W3134581569 cites W3108878197 @default.
• W3134581569 cites W3122224809 @default.
• W3134581569 cites W3123693939 @default.
• W3134581569 cites W3157608626 @default.
• W3134581569 cites W607505555 @default.
• W3134581569 cites W87140126 @default.
• W3134581569 cites W88685657 @default.
• W3134581569 cites W2465748195 @default.
• W3134581569 hasPublicationYear "2021" @default.
• W3134581569 type Work @default.
• W3134581569 sameAs 3134581569 @default.
• W3134581569 citedByCount "5" @default.
• W3134581569 countsByYear W31345815692021 @default.
• W3134581569 crossrefType "posted-content" @default.
• W3134581569 hasAuthorship W3134581569A5000011776 @default.
• W3134581569 hasAuthorship W3134581569A5005455141 @default.
• W3134581569 hasAuthorship W3134581569A5068556649 @default.
• W3134581569 hasAuthorship W3134581569A5080575294 @default.
• W3134581569 hasConcept C107321475 @default.
• W3134581569 hasConcept C111472728 @default.
• W3134581569 hasConcept C112680207 @default.
• W3134581569 hasConcept C112972136 @default.
• W3134581569 hasConcept C119857082 @default.
• W3134581569 hasConcept C124101348 @default.
• W3134581569 hasConcept C126255220 @default.
• W3134581569 hasConcept C12713177 @default.
• W3134581569 hasConcept C136764020 @default.
• W3134581569 hasConcept C138885662 @default.
• W3134581569 hasConcept C149728462 @default.
• W3134581569 hasConcept C154945302 @default.

• next