SVD Calculator

Calculate complete Singular Value Decomposition A = UΣVT

Matrix Input

Enter Matrix A

SVD Decomposition: A = UΣVT

Singular Values

σ1 = 6.3246
σ2 = 3.1623

Matrix U (2×2)

-0.4472
-0.8944
-0.8944
0.4472

Matrix Σ (2×2)

6.3246
0
0
3.1623

Matrix VT (2×2)

-0.7071
0.7071
-0.7071
-0.7071

Verification: UΣVT

4
0
3
-5

This should match your original matrix A (within numerical precision)

Step-by-Step Example

Given Matrix A

Consider the 2×2 matrix:

A = [4 0]
    [3 -5]

SVD Steps

Step 1: Calculate ATA

Step 2: Find eigenvalues of ATA

Step 3: Singular values σᵢ = √(eigenvalues)

Step 4: Form Σ matrix with singular values

Step 5: Calculate V from eigenvectors of ATA

Step 6: Calculate U from A·V = U·Σ

SVD Properties

U

Orthogonal Matrix

UTU = I

Left singular vectors

Σ

Diagonal Matrix

σ₁ ≥ σ₂ ≥ ... ≥ 0

Singular values

V

Orthogonal Matrix

VVT = I

Right singular vectors

SVD Applications

Data compression and noise reduction

Principal Component Analysis (PCA)

Least squares solutions

Image processing and recognition

Pseudoinverse computation

Recommendation systems

Understanding Singular Value Decomposition

What is SVD?

Singular Value Decomposition (SVD) is a fundamental matrix factorization that decomposes any real matrix A into the product of three matrices: A = UΣVT, where U and V are orthogonal matrices and Σ is diagonal.

Mathematical Foundation

  • A: Original m×n matrix
  • U: m×m orthogonal matrix (left singular vectors)
  • Σ: m×n diagonal matrix (singular values)
  • VT: n×n orthogonal matrix (right singular vectors)

Calculation Algorithm

Step 1: Compute ATA

Step 2: Find eigenvalues and eigenvectors of ATA

Step 3: Singular values σᵢ = √(eigenvalues)

Step 4: V matrix from eigenvectors of ATA

Step 5: U matrix from A·V = U·Σ

Key Properties

  • Exists for any real matrix (rectangular or square)
  • Singular values are non-negative and ordered
  • Optimal low-rank approximation
  • Reveals matrix rank and null space