Section 11: PCA

Jupyter Notebook Artur Rego Costa (2021), Danylo Lavrentovich (2020), adapted from Dan Greenfield, Verena Volf, and James Xue from past years

This Friday in section we will try and work on our general understanding of PCA, as well as give practical pointers for interpreting its output.

The plan: 1. hear the story of the one-dimensional scientist; 2. learn about the generative model of PCA with the goal of connecting it back to previous weeks' content; 3. go through a Jupyter notebook [download][view]. where we use PCA to analyze handwritten digits, with surprising similarities to this week's PSet!