Course contentsShow
← Mathematics
Lesson 2273 of 3,12147. Linear Algebra — Applications and Advanced TopicsFree lesson

Spectral Decomposition

Express a symmetric matrix as a sum of rank-one matrices using eigenvalues and orthonormal eigenvectors.

Spectral Decomposition

What You'll Learn

You'll learn how to express any symmetric matrix as a concrete sum of simpler rank-one matrices built from its eigenvalues and eigenvectors.

The Core Idea

The Spectral Theorem told us that every symmetric matrix can be orthogonally diagonalized: A = QDQ^T, where Q has orthonormal eigenvectors as columns and D is diagonal with eigenvalues.

Spectral Decomposition takes this one step further by expanding this product into an explicit sum. Instead of seeing A as one matrix product, we break it into pieces—each piece corresponding to one eigenvalue-eigenvector pair.

If A is an n×n symmetric matrix with eigenvalues λ₁, λ₂, ..., λₙ and corresponding orthonormal eigenvectors u₁, u₂, ..., uₙ, then:

A = λ₁u₁u₁ᵀ + λ₂u₂u₂ᵀ + ... + λₙuₙuₙᵀ

Each term λᵢuᵢuᵢᵀ is a rank-one matrix (it's an outer product of a vector with itself, scaled by the eigenvalue). Think of it as A being built from simple building blocks, each pointing in one eigenvector direction with strength given by the corresponding eigenvalue.

Why This Matters

This decomposition reveals the "spectrum" of A—how it's composed of independent directions. Each rank-one piece operates only along one eigenvector direction. When you multiply A by any vector, each piece contributes independently, stretching the component along uᵢ by the factor λᵢ.

It's like decomposing white light through a prism: the single matrix A splits into pure "colors" (eigenspaces), each with its own intensity (eigenvalue).

Key Takeaway: Spectral decomposition expresses a symmetric matrix as A = Σλᵢuᵢuᵢᵀ, decomposing it into a sum of rank-one matrices—one per eigenvalue-eigenvector pair—revealing its fundamental directional structure.