autoencoder (version 1.1)

autoencoder_Ninput=100_Nhidden=100_rho=1e-2:

A trained autoencoder example with 100 hidden units

Description

autoencoder.object is an example object of class autoencoder containing the weights, biases and other parameter of a sparse autoencoder with N.input=100, N.hidden=100, with sparsity parameter rho=0.01, trained on a dataset of 5000 image patches of 10 by 10 pixels, randomly cropped from decoloured nature photos.

Usage

data('autoencoder_Ninput=100_Nhidden=100_rho=1e-2')

Format

The format is: chr "autoencoder_Ninput=100_Nhidden=100_rho=1e-2"

Examples

## Load a pre-trained autoencoder object with N.input=100 and N.hidden=10*10,
## trained on unlabeled set of 5000 image patches of size Nx.patch by Ny.patch, 
## randomly picked from 10 nature photos, and visualize the features 
## learned by its hidden units:

data('autoencoder_Ninput=100_Nhidden=100_rho=1e-2')

## Visualize hidden units' learned features:
visualize.hidden.units(autoencoder.object,Nx.patch=10,Ny.patch=10)