Principal Component Analysis

Heard from CS231N, it gave these links: (wiki refCS229refblog ref)

From AlexNet paper, PCA color augmentation.

This seems like a really important topic for actual ML.

https://www.youtube.com/watch?v=FgakZw6K1QQ&ab_channel=StatQuestwithJoshStarmer

PCA is simply a method to reduce the dimensionality of data. You can see all about it here: https://www.askpython.com/python/examples/principal-component-analysis-for-image-data

It’s really cool to visualize higher dimensional Data. I see that people do PCA on higher dimensional data, before running K-Means Clustering, as can be seen in this official K-Means Clustering, as can be seen in this official scikit-learn tutorial