Projection and information loss
Read this example in chapter 14 →
Projection and information loss
Rotate a direction or keep fewer principal components. Which information survives?
Synthetic chapter 14 examples. Projection uses x = (3, 1). PCA uses the four centered observations from the book; its axes are fixed for this dataset, with 80% and 20% of the variance. This is not a general PCA solver.
Your changes stay in this page. Nothing is uploaded or saved. Reloading restores the original example.