mlearnx / mlseminars-autoencoders

Jupyter notebook of my autoencoder presentation

mlseminars-autoencoders

Slides for my autoencoder reading group seminar. Code based heavily on the keras library examples and keras blog (blog.keras.io).

Notebook can be viewed in github by clicking on Autoencoders.ipynb

or if you just want to look at the slide version: http://www.birving.com/presentations/autoencoders/index.html

How to run these slides and examples yourself

Setup python environment

• Jupyter notebook
• Requirements: numpy, keras, theano, jupyter
• Python 3
• Probably Python 2 as well but I haven't bothered to test -- everyone should be using python 3 ;)
• Install RISE for an interactive presentation viewer
  • see install instructions here (https://github.com/damianavila/RISE)

Get

# Clone the repository
git clone https://github.com/benjaminirving/mlseminars-autoencoders
cd mlseminars-autoencoders

Run

jupyter notebook

View in presentation mode using RISE
After running the notebook, click on the RISE button

or

jupyter nbconvert --to slides Autoencoders.ipynb --post serve

(for a non interactive version)