Essential Math for Data Science: Visual Introduction to Singular Value Decomposition - KDnuggets

In this article, you'll learn about Singular value decomposition (SVD), which is a major topic of linear algebra, data science, and machine learning. It is for instance used to calculate the Principal Component Analysis (PCA). You'll need some understanding of linear algebra basics (feel free to check the previous article and the book Essential Math for Data Science. You can only apply eigendecomposition to square matrices because it uses a single change of basis matrix, which implies that the initial vector and the transformed vector are relative to the same basis. You go to another basis with to do the transformation, and you come back to the initial basis with .