Sparse Zero-Sum Games as Stable Functional Feature Selection

Abstract : In large-scale systems biology applications, features are structured in hidden functional categories whose predictive power is identical. Feature selection, therefore, can lead not only to a problem with a reduced dimensionality, but also reveal some knowledge on functional classes of variables. In this contribution, we propose a framework based on a sparse zero-sum game which performs a stable functional feature selection. In particular, the approach is based on feature subsets ranking by a thresholding stochastic bandit. We provide a theoretical analysis of the introduced algorithm. We illustrate by experiments on both synthetic and real complex data that the proposed method is competitive from the predictive and stability viewpoints.
Document type :
Journal articles
Complete list of metadatas

Cited literature [38 references]  Display  Hide  Download

https://hal.sorbonne-universite.fr/hal-01223887
Contributor : Gestionnaire Hal-Upmc <>
Submitted on : Tuesday, November 3, 2015 - 4:02:23 PM
Last modification on : Friday, May 24, 2019 - 4:10:01 PM
Long-term archiving on : Thursday, February 4, 2016 - 11:16:49 AM

File

journal.pone.0134683.pdf
Publication funded by an institution

Licence


Distributed under a Creative Commons Attribution 4.0 International License

Identifiers

Citation

Nataliya Sokolovska, Olivier Teytaud, Salwa Rizkalla, Karine Clément, Jean-Daniel Zucker. Sparse Zero-Sum Games as Stable Functional Feature Selection. PLoS ONE, Public Library of Science, 2015, 10 (9), pp.e0134683. ⟨10.1371/journal.pone.0134683⟩. ⟨hal-01223887⟩

Share

Metrics

Record views

834

Files downloads

226