This is a practical part of a project created by Coursera Project Network. It introduces the theory behind an autoencoder (AE), its uses, and its advantages over PCA, a common dimensionality reduction technique.
- Generate and preprocess high-dimensional data
- Use cleaned data to create a PCA baseline model
- How an autoencoder works
- How to train an autoencoder in scikit-learn
- How to extract encoder from trained autoencoder
- How to evaluate dimensionality reduction using visual and analytical approaches