Unsupervised discovery of nonlinear structure using contrastive backpropagation.
We describe a way of modeling high-dimensional data vectors by using an unsupervised, nonlinear, multilayer neural network in which the activity of each neuron-like unit makes an additive contribution to a global energy score that indicates how surprised the network is by the data vector. The connec...
Главные авторы: | Hinton, G, Osindero, S, Welling, M, Teh, Y |
---|---|
Формат: | Journal article |
Язык: | English |
Опубликовано: |
2006
|
Схожие документы
-
Energy-based models for sparse overcomplete representations
по: Teh, Y, и др.
Опубликовано: (2004) -
A fast learning algorithm for deep belief nets.
по: Hinton, G, и др.
Опубликовано: (2006) -
Unsupervised part discovery from contrastive reconstruction
по: Choudhury, S, и др.
Опубликовано: (2021) -
Backpropagation and the brain
по: Lillicrap, TP, и др.
Опубликовано: (2020) -
Unsupervised discovery of parts, structure, and dynamics
по: Xu, Zhenjia, и др.
Опубликовано: (2020)