A supervised strategy for deep kernel machine

Abstract : This paper presents an alternative to the supervised KPCA based approach for learning a Multilayer Kernel Machine (MKM) . In our proposed procedure, the hidden layers are learnt in a supervised fashion based on kernel partial least squares regression. The main interest resides in a simplified learning scheme as the obtained hidden features are automatically ranked according to their correlation with the target outputs. The approach is illustrated on small scale real world applications and shows compelling evidences.
Document type :
Conference papers
Liste complète des métadonnées

Cited literature [9 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-00668302
Contributor : Maxime Berar <>
Submitted on : Monday, February 13, 2012 - 11:40:48 AM
Last modification on : Thursday, February 7, 2019 - 5:37:46 PM
Document(s) archivé(s) le : Wednesday, December 14, 2016 - 4:55:37 AM

File

es2011-21.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-00668302, version 1

Citation

Florian Yger, Maxime Berar, Gilles Gasso, Alain Rakotomamonjy. A supervised strategy for deep kernel machine. European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, Apr 2011, Bruges, Belgium. pp.501-506. ⟨hal-00668302⟩

Share

Metrics

Record views

205

Files downloads

213