SVD in PCA: How Singular Value Decomposition Powers Principal Component Analysis
Imagine you have a dataset—maybe a table with different measurements like height and weight of people. Some of these measurements might be highly related (e.g., height and weight are often correlated). Instead of storing all this information separately, wouldn’t it be nice to find a smaller set of "essential features" that still capture most of the original information?
Interestingly, PCA is computed using SVD! That means SVD is the mathematical engine behind PCA.
SVD Breaks Down Any Matrix
