其中U是 AAT 之eigenvectors;V是 ATA 之eigenvectors
4. For any matrix A, AAT and ATA are both symmetric.
sol. (AAT)T=AAT; (ATA)T=ATA
5. For arbitrary matrices A, B, AB and BA have same nonzero eigenvalues. (if AB, BA both exist)
sol. AB→u=λ→u→sinceλ,→u≠0,B→u≠0 and
BA(B→u)(→set=→y)=B(AB→u)=λB→u=λ→y
{AB→u=λ→uBA→y=λ→y
{AB→u=λ→uBA→y=λ→y
From 4. and 5., we can write ATA and AAT as:
AAT=UDUT; ATA=V˜DV and D, ˜D have the same nonzero eigenvalues.
6. Eigenvalues of AAT are all ≥ 0
sol. Let →v be an arbitrary vector with shape (A.shape[0], 1).
→vT1×mAm×nATn×m→vm×1=→yT→y≥0 (set y=→vTA)
Now we set →v= eigenvector of AAT (→u)
→uTAAT→u=→uTλ→u=λ→uT→u
Since →uT→u≥0 and →uTAAT→u≥0, λ always ≥0
And from 6. we can rewrite A as USVT such that
AAT=USVTVSTUT=UDUT and ATA=VSTUTUSVT=V˜DVT
Si,i的值 =
{√λi必存在 (for nonzero eigenvalue λi)0
沒有留言:
張貼留言