Course contentsShow
← Machine Learning and Deep Learning
Lesson 17 of 3,5381. Mathematical Foundations for Machine LearningFree lesson

Computing Eigenvalues and Eigenvectors

Learn the characteristic equation method and computational approaches for finding eigenvalues and eigenvectors.

Computing Eigenvalues and Eigenvectors

You'll learn how to actually find the eigenvalues and eigenvectors of a matrix using the characteristic equation.

The Core Idea

In the previous lesson, you learned what eigenvalues and eigenvectors are—special scalars and directions where a matrix just stretches or shrinks vectors. Now we need to find them practically.

The characteristic equation is our main tool. Remember that for an eigenvalue λ and eigenvector v, we have:

A****v = λv

Rearranging: A****v - λv = 0

This becomes: (A - λI)v = 0

For a non-zero eigenvector to exist, the matrix (A - λI) must be singular (non-invertible). From our determinant lesson, we know this means its determinant equals zero:

det(A - λI) = 0

This is the characteristic equation—a polynomial in λ whose solutions are the eigenvalues!

The Step-by-Step Process

  1. Find eigenvalues: Compute det(A - λI) and solve for λ
  2. Find eigenvectors: For each eigenvalue λ, solve (A - λI)v = 0 (this is a null space problem!)

Simple Example (2×2 case)

For matrix A = [[3, 1], [1, 3]]:

  • Compute det([[3-λ, 1], [1, 3-λ]]) = (3-λ)² - 1 = λ² - 6λ + 8 = 0
  • Solve: λ = 4 or λ = 2 (our eigenvalues!)
  • For λ=4: solve (A - 4I)v = 0 to get the eigenvector

Key Takeaway: Computing eigenvalues means solving the characteristic equation det(A - λI) = 0, then finding eigenvectors by solving the null space of (A - λI) for each eigenvalue. This connects determinants, null spaces, and linear systems—concepts you've already mastered!