Skip to content
Sahithyan's S1
Sahithyan's S1 — Mathematics

Eigenvalues & Eigenvectors

Definitions

Characteristic Polynomial

Let AA be a n×nn\times n matrix.

p(λ)=AλIp(\lambda)=|A-\lambda I|

Eigenvalues

Roots of the equation p(λ)=0p(\lambda) = 0 are the eigenvalues of AA.

Eigenvectors

The column vectors satisfying the equation (AλiI)Xi(A-\lambda_i I)X_i.

Normalized eigenvectors

An eigenvector with the magnitude (norm) of 11. Normalizing factor kk of any eigenvector is:

1k=i=1nXi2\frac{1}{k}=\sqrt{\sum_{i=1}^n X_i^2}

Norm

Norm of a column or row matrix Wn×nW_{n\times n} is denoted by W||W|| and defined as:

W=i=1nwi2||W|| = \sqrt{\sum_{i=1}^n w_i^2}

Algebraic Multiplicity

If the characteristic polynomial consists of a factor of the form (λλi)r(\lambda − \lambda_i)^r and (λλi)r+1(\lambda − \lambda_i)^{r+1} is not a factor of the characteristic polynomial then rr is the algebraic multiplicity of the eigenvalue λ\lambda.

Spectrum

Set of all eigenvalues.

Spectral Radius

R=max{λi  where  λiSpectrum}R=\max\Big\{|\lambda_i|\; where \;\lambda_i \in \text{Spectrum}\Big\}

Linear Independence of Eigenvectors

Suppose X1,X2,X3,,XnX_1,X_2,X_3,\dots,X_n is a set of eigenvectors. k1,k2,k3,,knk_1,k_2,k_3,\dots,k_n is a set of scalars.

All those eigenvectors are independent iff:

k1X1+k2X2+k3X3++knXn=0    k1=k2=k3==kn=0k_1X_1+k_2X_2+k_3X_3+\dots+k_nX_n=0 \implies k_1=k_2=k_3=\dots=k_n=0

For special matrices

Real symmetric matrix

Suppose AA is a symmetric matrix with all real entries. Then:

  • The eigenvalues of AA are all real: λSA,(λiR)\forall \lambda \in S_A, (\lambda_i \in \mathbb{R})
  • The eigenvectors of AA (corresponding to distinct values of λ\lambda) are mutually orthogonal
  • AA and ATA^T have the same eigenvalues

Upper triangular matrix

The eigenvalues are the diagonal entries.