https://machine-learning-course.readthedocs.io/en/latest/content/unsupervised/pca.html

University of Michigan - Ann Arbor

The attached image illustrates what PCA accomplishes on a two-dimensional scatterplot. The first principal component is chosen so that it captures the largest possible amount of variation in the data; here it runs from the bottom-left to the top-right of the plot, aligning with the direction of greatest spread. The second principal component is then set in a direction orthogonal to the first that captures the next most variation, explaining the spread not accounted for by the first component.

Learn Before