Unsupervised Learning

Principal Component Analysis (PCA)

• Read Chapter 10.1-10.2.1 in the book Introduction to Statistical Learning Links to an external site.

Clustering

• Read Chapter 10.3 in the book Introduction to Statistical Learning Links to an external site.
• For an overview of clustering with scikit-learn quickly also glance through (but don't get stuck) this overview http://scikit-learn.org/stable/modules/clustering.html Links to an external site.
• K-means clustering interactive demo Links to an external site.

Word2Vec

Word2vec is a way to find a vector space representation for words given a text corpus. 

Word2vec description and fun observations Links to an external site.

Explore word analogies, word2vec illustration Links to an external site.

Learning to predict the next word [Neural Networks for Machine Learning] Links to an external site.

Word2vec tutorial in Tensorflow Links to an external site.

T-SNE

T-sne is a method to visualize high-dimensional data in a way that tries to illustrate inherent structure.

https://en.wikipedia.org/wiki/T-distributed_stochastic_neighbor_embedding Links to an external site.

Google Tech Talk about t-SNE

Autoencoders

Auto-encoders can be used to automatically find low-dimensional features that capture structure in your dataset.

Check out this tutorial Links to an external site.. (You don't need to run the code) or this video Links to an external site.

Here is a nice example of use Links to an external site.