« Extinction de neurone » : différence entre les versions


mAucun résumé des modifications
Balise : Éditeur de wikicode 2017
(Ajustement de quelques sections de la fiche.)
Balise : Éditeur de wikicode 2017
Ligne 1 : Ligne 1 :
==Domaine==
==Domaine==
[[Category:Vocabulary]]<br/> [[Category:Intelligence artificielle]]Intelligence artificielle<br/>  [[Catégorie:Apprentissage automatique]] Apprentissage automatique<br/> [[Catégorie:Apprentissage profond]] Apprentissage profond<br/> [[Category:Coulombe]]Coulombe<br/>
[[Category:Vocabulaire]]<br/>  
[[Category:Intelligence artificielle]]Intelligence artificielle<br/>   
[[Category:Apprentissage automatique]] Apprentissage automatique<br/>  
[[Category:Apprentissage profond]] Apprentissage profond<br/>
[[Category:Coulombe]]<br/>
[[Category:Termino 2019]]
[[Category:Scotty]]
 


==Définition==
==Définition==
Méthode de régularisation de l'apprentissage d'un réseau profond par extinction de neurones (dropout). Rappelons que la régularisation consiste en l'ajout de contraintes pour réduire le surajustement (overfitting).<br/>
Technique de régularisation dont le principe est de désactiver aléatoirement à chaque itération un certain pourcentage des neurones d’une couche afin d'éviter le surajustement.
 
 
<br/>
<br/>


==Français==
==Français==
'''extinction de neurone'''


<poll>
<poll>
Ligne 22 : Ligne 32 :
==Anglais==
==Anglais==


===Dropout===
'''dropout'''


Dropout is a regularization technique for Neural Networks that prevents overfitting. It prevents neurons from co-adapting by randomly setting a fraction of them to 0 at each training iteration. Dropout can be interpreted in various ways, such as randomly sampling from an exponential number of different networks. Dropout layers first gained popularity through their use in CNNs, but have since been applied to other layers, including input embeddings or recurrent networks.
Dropout is a regularization technique for Neural Networks that prevents overfitting. It prevents neurons from co-adapting by randomly setting a fraction of them to 0 at each training iteration. Dropout can be interpreted in various ways, such as randomly sampling from an exponential number of different networks. Dropout layers first gained popularity through their use in CNNs, but have since been applied to other layers, including input embeddings or recurrent networks.
*Dropout: A Simple Way to Prevent Neural Networks from Overfitting
*Recurrent Neural Network Regularization


<br />
<br />

Version du 3 juin 2019 à 19:30

Domaine


Intelligence artificielle
Apprentissage automatique
Apprentissage profond


Définition

Technique de régularisation dont le principe est de désactiver aléatoirement à chaque itération un certain pourcentage des neurones d’une couche afin d'éviter le surajustement.



Français

extinction de neurone

<poll> Choisissez parmi ces termes proposés : extinction de neurone maille de réseau point mort trou de neurone trou de mémoire trou de réseau </poll>

Anglais

dropout

Dropout is a regularization technique for Neural Networks that prevents overfitting. It prevents neurons from co-adapting by randomly setting a fraction of them to 0 at each training iteration. Dropout can be interpreted in various ways, such as randomly sampling from an exponential number of different networks. Dropout layers first gained popularity through their use in CNNs, but have since been applied to other layers, including input embeddings or recurrent networks.