RFN-package: Rectified Factor Networks

In bioinf-jku/librfn-R: Rectified Factor Networks

Description

This package implements the Rectified Factor Networks algorithm as described by Clevert et al., 2014, which is an unsupervised latent variable model.

Details

Package:	RFN
Type:	Package
Version:	0.1
Date:	2016-02-19
License:	GPL (>= 2)

The core of this package is the function train_rfn, which trains a Rectified Factor Network as described by Clevert et al., 2014. Alongside, there exist three basic functions for plotting data, particularly MNIST pictures, implemented in showSinglePicture, showSingleColorPicture and showPictures.

Author(s)

Thomas Unterthiner, Thomas Adler

Maintainer: Thomas Unterthiner <[e-mail]>

References

Clevert DA, Unterthiner T, Mayr A, Ramsauer H, Hochreiter S, "Rectified Factor Networks", http://arxiv.org/abs/1502.06464

bioinf-jku/librfn-R documentation built on May 12, 2019, 9:24 p.m.