Day 22

Today

• • Clustering

  • Work on projects

For Next Class

• • Continue to work on projects

  • Next class topic: experimental design and control variables (no prereading)

Clustering

The context for today was the ipython notebook by Jake VanderPlas on the k-means algorithm.  This is an absolutely fantastic visual explanation of the various steps of the algorithm.  Today, we're going to expand on this in two important ways:

• 1. Dig a bit more into the theory of the k-means algorithm.

  2. Discuss some practical considerations for k-means.

The materials for today are located in the DataScience16 repo under day21.