ETH Zurich, Department of Architecture.
Streamed on April 17, 2018.
Vahid Moosavi - Machine Learning and Engineering Applications
S02E05- From PCA to AutoEncoders
Github Repo: https://github.com/sevamoo/data_driven_modeling_2018
Topics:
- Linear Transformations
- PCA
- Overcomplete representations: PCA in comparison to SOM, K-Means
- Centralized representation Vs. Distributed representation
- Auto-Encoders
- Deep Auto-Encoders
- Thinking about “computational graphs”