Overview

R250: Advanced Topics in Machine Learning is a course taken by MPhil ACS and Part III students. Students pick 3 topics, and present papers. This page describes the topic on Autoencoders.

• Briefing slides
• Introductory lecture slides and notebook with pytorch/mnist code
• Variational auto-encoder slides
• All course arrangements are on Moodle

Papers

1. Extracting and composing robust features with denoising autoencoders, Vincent, Larochelle, Bengio, Manzagol (ICML 2008)
2. Learning adversarially fair and transferable representations, Madras, Creager, Pitassi, Zemel (ICML 2018)
3. Auto-Encoding Variational Bayes, Kingma, Welling (ICLR 2013)
4. Semi-supervised learning with deep generative models, Kingma, Rezende, Mohamed, Welling (NIPS 2014)
5. β-VAE: learning basic visual concepts with a constrained variational framework, Higgins, Matthey, Pal, Burgess, Glorot, Botbinick, Mohamed, Lerchner (ICLR 2017)
6. Grammar Variational Autoencoder, Kisner, Paige, Hernández-Lobato (ICML 2017)