Communication Dans Un Congrès Année : 2017

Magnitude-Preserving Ranking for Structured Outputs

Résumé

In this paper, we present a novel method for solving structured prediction problems, based on combining Input Output Kernel Regression (IOKR) with an extension of magnitudepreserving ranking to structured output spaces. In particular, we concentrate on the case where a set of candidate outputs has been given, and the associated pre-image problem calls for ranking the set of candidate outputs. Our method, called magnitude-preserving IOKR, both aims to produce a good approximation of the output feature vectors, and to preserve the magnitude differences of the output features in the candidate sets. For the case where the candidate set does not contain corresponding 'correct' inputs, we propose a method for approximating the inputs through application of IOKR in the reverse direction. We apply our method to two learning problems: cross-lingual document retrieval and metabolite identi cation. Experiments show that the proposed approach improves performance over IOKR, and in the latter application obtains the current state-of-the-art accuracy.

Fichier principal
Vignette du fichier
brouard17a_1.pdf (468.46 Ko) Télécharger le fichier
Origine Fichiers éditeurs autorisés sur une archive ouverte
Licence
Loading...

Dates et versions

hal-02734693 , version 1 (02-06-2020)

Licence

Identifiants

  • HAL Id : hal-02734693 , version 1
  • PRODINRA : 476222

Citer

Celine Brouard, Eric Bach, Sebastian Böcker, Juho Rousu. Magnitude-Preserving Ranking for Structured Outputs. Asian conference on machine learning (ACML 2017), Nov 2017, Seoul, South Korea. 16 p. ⟨hal-02734693⟩

Collections

62 Consultations
105 Téléchargements

Partager

  • More