eigenvector/ eigenvalue
let A be a square nxn matrix. A non-zero n vector v is a eigenvector of A with eigenvalue λ if Av = λv
or (A-Iλ)v = 0
whens non-zero v
iff det(A - Iλ) = 0 which implies A - I λ non-invertible
characteristic polynomial
Pa(t) = det(A - tI)
Pa(λ) = 0 to find λ (eigenvalues)
algebraic multiplicity
P(t) = C(t-λ1)^k1(t-√2)^k2….(t-λp)^kp
∑ki = N , λi ≠ λj
λi has algebraic multiplicity ki
eigenspace
for eigenvalue λ with algebraic multiplicity k the eigenspace V is the P dimensional vector space spanned by these eigenvectors
Vλ = ker (A - Iλ)
(when A - Iλ = 0)
algepraic multiplicity k meaning
if λ has algebraic multiplicity k then there are P linearly independent eigenvectors where 1<= p<= k
geometric multiplicity
P = dimV is the geometric multiplicity of λ