« Machine de Helmholtz » : différence entre les versions
(Page créée avec « == en construction == == Définition == XXXXXX Voir aussi '''machine de Boltzmann''', '''Hopfield Network''' et '''réseau d'inférence''' == Français == ''' Machine de Helmholtz''' == Anglais == ''' Helmholtz Machine''' ''The Helmholtz machine is a type of artificial neural network that can account for the hidden structure of a set of data by being trained to create a generative model of the original set of data.'' == Source == [https://hilb... ») |
Aucun résumé des modifications |
||
| Ligne 12 : | Ligne 12 : | ||
''' Helmholtz Machine''' | ''' Helmholtz Machine''' | ||
'' | ''Helmholtz machine is a type of artificial neural network with a multilayer architecture trainable by an unsupervised learning algorithm such as the wake sleep algorithm. It learns first how to generate the incoming data for itself, in other words, to self-organize instead of classifying the data.'' | ||
''It contains two networks: a bottom-up recognition one and a top-down generation one. '' | |||
''The Helmholtz machine is a precursor to variational autoencoders.'' | |||
== Source == | == Source == | ||
Version du 15 novembre 2024 à 15:54
en construction
Définition
XXXXXX
Voir aussi machine de Boltzmann, Hopfield Network et réseau d'inférence
Français
Machine de Helmholtz
Anglais
Helmholtz Machine
Helmholtz machine is a type of artificial neural network with a multilayer architecture trainable by an unsupervised learning algorithm such as the wake sleep algorithm. It learns first how to generate the incoming data for itself, in other words, to self-organize instead of classifying the data.
It contains two networks: a bottom-up recognition one and a top-down generation one.
The Helmholtz machine is a precursor to variational autoencoders.
Source
Contributeurs: Arianne
