Singular Value Decomposition

Factor a matrix as U Σ Vᵀ

Matrix A

A = U Σ Vᵀ
Singular values
5, 3
Rank
2
Condition number
1.6667
Spectral norm (σ₁)
5
Frobenius norm
5.830952
Full rank
U — left singular vectors
0.7071
0.7071
0.7071
-0.7071
Σ — singular values
5
0
0
3
Vᵀ — right singular vectors
0.7071
0.7071
0
0.2357
-0.2357
0.9428

Multiplying U, Σ and Vᵀ back together reproduces the original matrix A.

Singular Value DecompositionThe unit circle is mapped by A to an ellipse with semi-axes 5 and 3, which are the two singular values.σ = 5σ = 3Unit circleImage A·x