The Singular Value Decomposition
![](https://velog.velcdn.com/images/dmg919/post/1000a3ed-7809-4c3c-9d38-382d1ffc2b22/image.png)
example
![](https://velog.velcdn.com/images/dmg919/post/6aaeb166-86a3-4315-8e48-6943425f6a61/image.png)
Theorem.
![](https://velog.velcdn.com/images/dmg919/post/2eaf2d37-c1cf-4769-9caa-1ae83dd552ef/image.png)
definition
![](https://velog.velcdn.com/images/dmg919/post/aa8ecc74-6787-46cd-a3c2-5273aab2ee60/image.png)
derivation
![](https://velog.velcdn.com/images/dmg919/post/c698faed-0b3c-40a7-84ea-55636f9a3303/image.png)
Bases for Fundamental Subspaces (+visualizing)
![](https://velog.velcdn.com/images/dmg919/post/5795e508-342b-46d8-88b7-2ad0d9b3d2ad/image.png)
Invertible matrix theorem (추가)
![](https://velog.velcdn.com/images/dmg919/post/06d68be3-ae46-4030-acbb-865caeb89bb0/image.png)
Reduced SVD
![](https://velog.velcdn.com/images/dmg919/post/892ebbf3-0b48-4b11-bbd9-3eb5563e8561/image.png)
pseudoinverse
![](https://velog.velcdn.com/images/dmg919/post/51110f99-501a-44d8-81a6-c4768ba0c776/image.png)
x̂은 Ax=b의 least-squares solution, smallest length
Application
- least-squares solution
- data analysis (covariance matrix)
- image / sound processing
- etc.