Apprentissage statistique en grande dimension et application au diagnostic oncologique par radiomique

Abstract : With the increase in measurement capabilities, many medical disciplines have seen their practices deeply modified because of the dimensionality of the data. Although these technical improvements promise significant advances in medical research, the statistical learning methods must be able to cope with the problems encountered in those high-dimensional spaces. The subspace classification and "sparse" methods introduced in recent years propose to meet this expectation. This article presents a quick overview of these difficulties and the proposed solutions, as well as an illustration of the use of one of these solutions for oncology diagnosis with radiomics.
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-01884468
Contributor : Charles Bouveyron <>
Submitted on : Monday, October 1, 2018 - 9:36:12 AM
Last modification on : Thursday, February 7, 2019 - 4:53:44 PM
Document(s) archivé(s) le : Wednesday, January 2, 2019 - 12:43:27 PM

File

Article-ANM.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01884468, version 1

Collections

Citation

Charles Bouveyron. Apprentissage statistique en grande dimension et application au diagnostic oncologique par radiomique. Cédric Villani; Bernard Nordlinger. Santé et intelligence artificielle, CNRS Editions, pp.179-189, 2018. ⟨hal-01884468⟩

Share

Metrics

Record views

112

Files downloads

116