Titre : |
L’approche neuronale de l’inférence statistique. |
Type de document : |
texte imprimé |
Auteurs : |
Dalel Zerdazi, Auteur ; Ahmed Chibat, Directeur de thèse |
Editeur : |
جامعة الإخوة منتوري قسنطينة |
Année de publication : |
2017 |
Importance : |
152 f. |
Format : |
30 cm. |
Note générale : |
2 copies imprimées disponibles
|
Langues : |
Français (fre) |
Catégories : |
Français - Anglais Mathématiques
|
Tags : |
Régression linéaire Régession ridge Estimateurs concurrents Modèles
linéaires généralisés Inférence sous contraintes Réseaux de neurones et approches connexes Apprentissage et systèmes adaptatifs Linear regression Ridge regression Shrinkage estimators Generalized linear models Inference under constraints Neural nets and related approaches Learning and adaptive systems |
Index. décimale : |
510 Mathématiques |
Résumé : |
This thesis is an attempt to contribute, even slightly, in situating the
neural networks theory into the framework of applied statistics. The central
issue of statistical inference was studied under the light of neural approach.
A lot of attention was payed to the notion of generalization, with the aim
to conceive an unified approach, ghattering toghether traditional statistical
methods with those resulting from neural networks theory, and presenting
them as emerging from the same principle.
The competitor estimators to the least squares one are surveyed, this is also
done for the different neural techniques conceived for the needs of regression
and prediction. A comparative study was done with the aim to show that
the fondamental concept, at the level of the roots, of these different methods
can be seen as unique.
An application case is presented to illustrate the predictive power of neural
nets versus the classical methods. |
Diplôme : |
Doctorat en sciences |
En ligne : |
../theses/math/ZER7034.pdf |
Format de la ressource électronique : |
pdf |
Permalink : |
index.php?lvl=notice_display&id=10565 |
|