Matches in SemOpenAlex for { <https://semopenalex.org/work/W3171223447> ?p ?o ?g. }
- W3171223447 abstract "Extreme multi-label classification (XMC) aims to learn a model that can tag data points with a subset of relevant labels from an extremely large label set. Real world e-commerce applications like personalized recommendations and product advertising can be formulated as XMC problems, where the objective is to predict for a user a small subset of items from a catalog of several million products. For such applications, a common approach is to organize these labels into a tree, enabling training and inference times that are logarithmic in the number of labels. While training a model once a label tree is available is well studied, designing the structure of the tree is a difficult task that is not yet well understood, and can dramatically impact both model latency and statistical performance. Existing approaches to tree construction fall at an extreme point, either optimizing exclusively for statistical performance, or for latency. We propose an efficient information theory inspired algorithm to construct intermediary operating points that trade off between the benefits of both. Our algorithm enables interpolation between these objectives, which was not previously possible. We corroborate our theoretical analysis with numerical results, showing that on the Wiki-500K benchmark dataset our method can reduce a proxy for expected latency by up to 28% while maintaining the same accuracy as Parabel. On several datasets derived from e-commerce customer logs, our modified label tree is able to improve this expected latency metric by up to 20% while maintaining the same accuracy. Finally, we discuss challenges in realizing these latency improvements in deployed models." @default.
- W3171223447 created "2021-06-22" @default.
- W3171223447 creator A5028381995 @default.
- W3171223447 creator A5040127066 @default.
- W3171223447 creator A5051353645 @default.
- W3171223447 creator A5055644525 @default.
- W3171223447 creator A5063459703 @default.
- W3171223447 date "2021-06-01" @default.
- W3171223447 modified "2023-09-27" @default.
- W3171223447 title "Enabling Efficiency-Precision Trade-offs for Label Trees in Extreme Classification" @default.
- W3171223447 cites W1485052049 @default.
- W3171223447 cites W1661372310 @default.
- W3171223447 cites W1723811852 @default.
- W3171223447 cites W1834987204 @default.
- W3171223447 cites W2060108852 @default.
- W3171223447 cites W2061873838 @default.
- W3171223447 cites W2068074736 @default.
- W3171223447 cites W2131462252 @default.
- W3171223447 cites W2153579005 @default.
- W3171223447 cites W2169054943 @default.
- W3171223447 cites W2437817353 @default.
- W3171223447 cites W2468328197 @default.
- W3171223447 cites W2520348554 @default.
- W3171223447 cites W2739996966 @default.
- W3171223447 cites W2743021690 @default.
- W3171223447 cites W2782759081 @default.
- W3171223447 cites W2788125153 @default.
- W3171223447 cites W2886971574 @default.
- W3171223447 cites W2897031196 @default.
- W3171223447 cites W2947118660 @default.
- W3171223447 cites W2950577311 @default.
- W3171223447 cites W2987098737 @default.
- W3171223447 cites W2992422826 @default.
- W3171223447 cites W3080739421 @default.
- W3171223447 cites W3080802002 @default.
- W3171223447 cites W3093858897 @default.
- W3171223447 cites W36903255 @default.
- W3171223447 cites W756166754 @default.
- W3171223447 cites W2752870707 @default.
- W3171223447 hasPublicationYear "2021" @default.
- W3171223447 type Work @default.
- W3171223447 sameAs 3171223447 @default.
- W3171223447 citedByCount "0" @default.
- W3171223447 crossrefType "posted-content" @default.
- W3171223447 hasAuthorship W3171223447A5028381995 @default.
- W3171223447 hasAuthorship W3171223447A5040127066 @default.
- W3171223447 hasAuthorship W3171223447A5051353645 @default.
- W3171223447 hasAuthorship W3171223447A5055644525 @default.
- W3171223447 hasAuthorship W3171223447A5063459703 @default.
- W3171223447 hasConcept C113174947 @default.
- W3171223447 hasConcept C119857082 @default.
- W3171223447 hasConcept C124101348 @default.
- W3171223447 hasConcept C13280743 @default.
- W3171223447 hasConcept C134306372 @default.
- W3171223447 hasConcept C154945302 @default.
- W3171223447 hasConcept C177264268 @default.
- W3171223447 hasConcept C185798385 @default.
- W3171223447 hasConcept C199360897 @default.
- W3171223447 hasConcept C205649164 @default.
- W3171223447 hasConcept C2776214188 @default.
- W3171223447 hasConcept C33923547 @default.
- W3171223447 hasConcept C41008148 @default.
- W3171223447 hasConcept C76155785 @default.
- W3171223447 hasConcept C82876162 @default.
- W3171223447 hasConceptScore W3171223447C113174947 @default.
- W3171223447 hasConceptScore W3171223447C119857082 @default.
- W3171223447 hasConceptScore W3171223447C124101348 @default.
- W3171223447 hasConceptScore W3171223447C13280743 @default.
- W3171223447 hasConceptScore W3171223447C134306372 @default.
- W3171223447 hasConceptScore W3171223447C154945302 @default.
- W3171223447 hasConceptScore W3171223447C177264268 @default.
- W3171223447 hasConceptScore W3171223447C185798385 @default.
- W3171223447 hasConceptScore W3171223447C199360897 @default.
- W3171223447 hasConceptScore W3171223447C205649164 @default.
- W3171223447 hasConceptScore W3171223447C2776214188 @default.
- W3171223447 hasConceptScore W3171223447C33923547 @default.
- W3171223447 hasConceptScore W3171223447C41008148 @default.
- W3171223447 hasConceptScore W3171223447C76155785 @default.
- W3171223447 hasConceptScore W3171223447C82876162 @default.
- W3171223447 hasLocation W31712234471 @default.
- W3171223447 hasOpenAccess W3171223447 @default.
- W3171223447 hasPrimaryLocation W31712234471 @default.
- W3171223447 hasRelatedWork W2060737402 @default.
- W3171223447 hasRelatedWork W2518508657 @default.
- W3171223447 hasRelatedWork W2550446025 @default.
- W3171223447 hasRelatedWork W2591575955 @default.
- W3171223447 hasRelatedWork W2605289741 @default.
- W3171223447 hasRelatedWork W2739621268 @default.
- W3171223447 hasRelatedWork W2766377147 @default.
- W3171223447 hasRelatedWork W2799563188 @default.
- W3171223447 hasRelatedWork W2803823445 @default.
- W3171223447 hasRelatedWork W2953280125 @default.
- W3171223447 hasRelatedWork W3013283295 @default.
- W3171223447 hasRelatedWork W3026606204 @default.
- W3171223447 hasRelatedWork W3037308981 @default.
- W3171223447 hasRelatedWork W3091262782 @default.
- W3171223447 hasRelatedWork W3097254552 @default.
- W3171223447 hasRelatedWork W3172459354 @default.
- W3171223447 hasRelatedWork W3189632814 @default.
- W3171223447 hasRelatedWork W3191265128 @default.