Manifold Structured Prediction

Structured prediction provides a general framework to deal with supervised problems where the outputs have semantically rich structure. While classical approaches consider finite, albeit potentially huge, output spaces, in this paper we discuss how structured prediction can be extended to a continuous scenario. Specifically, we study a structured prediction approach to manifold valued regression. We characterize a class of problems for which the considered approach is statistically consistent and study how geometric optimization can be used to compute the corresponding estimator. Promising experimental results on both simulated and real data complete our study.

Domains

Machine Learning [cs.LG]

Fichier principal

arxiv.pdf (418.26 Ko)

arxiv.log (60.91 Ko)

arxiv.out (1.53 Ko)

arxiv.synctex.gz (191.25 Ko)

pictures/expmap.png (43.45 Ko)

pictures/fprints.png (202.64 Ko)

pictures/learn_hess_FLOSS.png (34.68 Ko)

pictures/learn_hess_GLOSS.png (41.83 Ko)

pictures/learn_inv_FLOSS.png (37.07 Ko)

pictures/learn_inv_GLOSS.png (30.26 Ko)

pictures/partrans.png (38.4 Ko)

Origin	Files produced by the author(s)

Alessandro Rudi : Connect in order to contact the contributor

https://hal.science/hal-01958900

Submitted on : Wednesday, December 19, 2018-12:01:04 AM

Last modification on : Wednesday, December 18, 2024-9:04:10 AM

Dates and versions

hal-01958900 , version 1 (19-12-2018)

Identifiers

HAL Id : hal-01958900 , version 1
ARXIV : 1806.09908

Cite

Alessandro Rudi, Carlo Ciliberto, Gian Maria Marconi, Lorenzo Rosasco. Manifold Structured Prediction. NIPS 2018 - Neural Information Processing Systems Conference, Dec 2018, Montreal, Canada. pp.5615-5626. ⟨hal-01958900⟩

Export

BibTeX XML-TEI Dublin Core DC Terms EndNote DataCite

Collections

ENS-PARIS CNRS INRIA INRIA2 PSL

71 View

64 Download