Betreuer: Heiko Donat
Representation Learning describes the a set of techniques to extract features from raw data. In this seminar work, the general concept of autoencoders shall be described and techniques for of autoencoder-based representation learning will be summarized and explained.
This bachelor level seminar topic will be based on [1] and [2] with an additional literature research about autoencoders.
[1] http://www.deeplearningbook.org/contents/representation.html
[2] Recent Advances in Autoencoder-Based Representation Learning