Back to the lessonPreparing the printout…

Mathematical Foundations for Data Science · Part I — Linear Algebra

Topic 4 · Eigenvalues and Eigenvectors

[video] Lecture video — Topic 4: Eigenvalues and Eigenvectors — /api/lesson-media/videos/4d3653ff-a3ca-4f63-bf8e-bb93c553e599-1794885371253-d76e9db608dc668d.mp4

Flashcards · Eigenvalues and eigenvectors

Eigenvector and eigenvalue — A nonzero v with Av = λv. Nothing else is required: λ may be 0 and v may have zero entries and any length.

Characteristic equation — det(A − λI) = 0. Its roots are the eigenvalues; for each, solve (A − λI)v = 0 for the eigenvectors.

A = [[4, 1], [2, 3]]: λ² − 7λ + 10 = 0, so λ = 2 or 5.

Diagonal and triangular matrices — The eigenvalues are the diagonal entries.

Trace and determinant — Sum of eigenvalues = trace; product of eigenvalues = det A. Use them as a check.

Inverse and powers — If Av = λv then Aᵏv = λᵏv, p(A)v = p(λ)v and, for invertible A, A⁻¹v = λ⁻¹v.

Diagonalizable — A = PDP⁻¹ with P invertible and D diagonal. Equivalent: A has n linearly independent eigenvectors.

Building P and D — Columns of P are eigenvectors; D has the matching eigenvalues in the same order (AP = PD).

Ghi chú: Column order and eigenvalue order must agree.

When it fails — Repeated eigenvalues may give too few eigenvectors. [[2, 1], [0, 2]] has one eigenvector line and is not diagonalizable; 2I is.

Powers via diagonalization — Aᵏ = PDᵏP⁻¹ with Dᵏ = diag(λ₁ᵏ, …, λₙᵏ).

[markdown] content available only on the platform

Mathematical Foundations for Data Science · Topic 4 · Eigenvalues and Eigenvectors · printed on November 20, 2026