Introducing complex dependency structures into supervised components-based models
Une forte redondance des variables explicatives cause de gros problèmes d'identifiabilité et d'instabilité des coefficients dans les modèles de régression. Même lorsque l'estimation est possible, l'interprétation des résultats est donc extrêmement délicate. Il est alors indispens...
Main Author: | Chauvet, Jocelyn |
---|---|
Other Authors: | Montpellier |
Language: | en |
Published: |
2019
|
Subjects: | |
Online Access: | http://www.theses.fr/2019MONTS008/document |
Similar Items
-
廣義線性混合模式結合B-Spline在疾病地圖上之應用
by: 連家斌 -
A framework for studying behavioral evolution by reconstructing ancestral repertoires
by: Damián G Hernández, et al.
Published: (2021-09-01) -
Relationship between the presence of Helicobacter pylori glmM gene and serum fluctuations of IgG and IgA antibodies in infected individuals
by: Ayat Moradipour, et al.
Published: (2021-01-01) -
Understanding correlates of child stunting in Ethiopia using generalized linear mixed models
by: Kasahun Takele, et al.
Published: (2019-05-01) -
Choosing priors in Bayesian ecological models by simulating from the prior predictive distribution
by: Jeff S. Wesner, et al.
Published: (2021-09-01)